The Chronology of the Reign of

Herod the Great


Journal of Theological Studies

(pp. 283-298)


W.E. Filmer



[Editor's Note: Filmer's work re-opened the issue of the year of Christ's birth by demonstrating the impossibility that Herod died in 4 B.C., which had previously been widely supposed.  Filmer's conclusion that Herod died in 1 B.C. is correct, but his conclusion that Herod died in the month of January probably cannot reconciled with scripture.  Luke's account states that the holy family returned to Nazareth after performing the customary rites at the temple forty days after Jesus' birth (Lk. 2:22, 39; Ex. 13:2; Lev. 12).  Assuming Christ was born Casleu 25, 2 B.C., forty days would place the holy family in Jerusalem Shebat 6 on the Jewish calendar, or February 3, on the Julian. The fact that the holy family departed from Bethlehem forty days after Christ's birth and then returned to Nazareth means that it was from Nazareth that Joseph fled to Egypt, and not from Bethlehem.  Thus, the better view is that the Magi arrived soon after the holy family's return to Nazareth, where they were guided by the Star, rather than Bethlehem where Herod sent them (Matt. 2:8-10). Roman records of the census would have recorded the presence of Joseph and Mary in Bethlehem and the birth of their child, and would have indicated that Nazareth was where they made their home.  Realizing that the child may not have remained in Bethlehem for two years, the earliest date of his birth based upon the Star that appeared to the Magi in the east, Herod would likely have consulted the Roman census to discover all male children born there during the relevant period.  The angel's warning that Herod would "seek the young child" (Matt. 2:13) indicates that Slaughter of the Innocents in Bethlehem would not exhaust his search, requiring Joseph to leave Herod's jurisdiction entirely.  As indicated by Josephus, Herod died shortly before Passover 1  B.C.]


For many years the dates of the birth and crucifixion of Christ have been matters of controversy. The birth of Christ must have been before the death of Herod the Great, for Herod, on hearing that a child had been born who was to become king of the Jews, ordered the massacre of all male children under two years of age in Bethlehem.   Now according to Josephus, Herod died shortly after an eclipse of t5eh moor and not long before a Passover.[1] Since there was an eclipse of the moon on the night of 12/13 March, 4 B.C., which was exactly a month before the Passover, it was been widely assumed that Herod died in the spring of that year, and consequently that Christ was born at least as early as 5 B.C. 

Within the possible limits for the date of the crucifixion, the 14th Nisan fell on a Friday only in the years A.D. 30 and 33,[2] and of these tow date recent opinion has tended to favour the latter.  But Luke 3:23 says that Jesus was “about thirty years of age” when he began his ministry, and this could not have been earlier than A.D. 29 if we take A.D. 33 for the crucifixion. Since Jesus may have been anything up to two years of age when Herod died, we find ourselves obliged either to accept Luke’s statement with an unduly large degree of latitude, or to question the evidence for the date of Herod’s death as early as 4 B.C. 

The Eclipse of the Moon 

Professor J. Finegan in his Handbook of Biblical Chronology, p. 231 (§ 365), quotes Emil Schurer as saying, “Only on the night of Mar. 12/13, 4 B.C. was there a lunar eclipse, and there was no such phenomenon in 3 or 2 B.C.  Accordingly the death of Herod took place between Mar 12 and Apr 11 in the year 4 B.C.  Reference to the English translation of Schurer’s work confirms that he did make such a statement, but in the German second and later editions he added, “Only in 5 B.C. on 15 Sept. and in 1 B.C. on 9 January did other lunar eclipses occur which were visible in Jerusalem.  But these cannot be considered on account of the other data.”[3]  Regarding the eclipse in 1 B.C.  he refers the reader to F. Riess, Das Geburiskjekr Christi (1880).  Reference to this work, however, reveals this Reiss believed that Herod did not die in 4 B.C. but soon after the eclipse in 1 B.C. because of the other data, namely the numerous events that took place between the eclipse and the Passover, count not be squeezed into the four weeks available in 4 B.C.  In other words, Reiss not only fails to support Schurer’s thesis, but flatly contradicts it. 

If Reiss’s arguments were completely valid, we should be obliged to reject 4 B.C. altogether; but he overstates his case[4] and although it is conceivable that all these events could have taken place in a month, it is nevertheless much more likely that two or three months elapsed between the eclipse and the Passover.  Three months are available in 1 B.C. and, moreover, the eclipse was total, as against only four digits on the earlier occasion.[5] 

However, the value of the lunar eclipse as a useful piece of evidence is further reduced when it is realized that yet another eclipse took place on the evening of 19 December, 1 B.C,[6] which fell three months before a Passover, that of A.D. 1.  This may have been overlooked, because its mid-point occurred shortly before the moon rose at Jerusalem; but since its magnitude was seven digits, it would still have been visible during the early evening when people would have likely to notice it. 

Thus, so far as the evidence of lunar eclipses goes, Herod may have died in either of the even 4 or 1 B.C. or even in A.D. 1. 

The Megillet Ta’anit 

Before going on to consider the historical background in more detail, it will be convenient at this state to mention the evidence provided by the Megillot Ta’anit, a Jewish list of days on which, by reason of notable events associated with them, the Jews were not allowed to fast.  The list was complied shortly before the destruction of the temple in A.D. 70, and the reason for the holiday is given in every case but two. S. Zeitlin explains that “undoubtedly the chronicler’s silence in these instances is due to their being recently instituted holidays pro tempore.  The incidents being well know to all, it was not necessary to add any explanations.”[7]  According to Jewish tradition these holidays, which fell on 7 Kislev and 2 Shebat, commemorated the death of Herod and the death of Jannai, both kings who had died within recent memory and whom the Jews hated. 

Now 7 Kislev, which the Jewish commentator actually preferred to associate with Herod, fell earlier in the year than any of the above three lunar eclipses, and for this reason it must be ruled out. But 2 Shebat, although it fell before the eclipse in 4 B.C., occurred fifteen days after eclipse in 1 B.C.  In view of the serious deterioration of Herod’s health which Josephus says set in immediately after the eclipse, it does not seem likely that he could have lived much longer than the fortnight that this allows. 

Herod’s Accession 

The principal historical evidence for the date of Herod’s death is provided by Josephus in two statements in which he says that he reigned thirty-four years from the death of his predecessor Antigonus, but thirty-seven years from the time when he was declare king by the Romans.[8]  It is therefore important to establish precisely these two dates for his accession.  Regarding his appointment in Rome, Josephus says, “Thus did this man receive the kingdom, having obtained it on the hundred and eighty-fourth Olympiad, when Gaius Domitius Calvinus was counsel the second time, and Caius Asinius Pollio he first time.”[9]  This makes f40 B.C. but in this he is contradicted by Appian, who mentions Herod’s appointment in a context that can be dated from Dio’s Roman History to 39 B.C.[10] 

Following his appointment, Herod had to fight three years for his kingdom against Antigonus who was ruling Jerusalem.  Finally, with the help of the Roman general Sosius, he took the city, says Josephus, “when Marcus Agrippa and Caninius Gallus were consuls at Rome, on the hundred and eighty-fifth Olympiad, on the third month, on the solemnity of the fast, as if a periodical revolution of calamities had returned, since that which befell the Jews under Pompey; for the Jews were taken by him on the same day, and this was after twenty-seven years times.”[11] 

The consular dating gives 37 B.C. but this does not agree with what follows: Pompey captured Jerusalem in 63 B.C., and twenty-seven years later would bring us to 36, not to 37 B.C.  The twenty-seven years might well be rejected if it stood alone, but Josephus states that on both occasions Jerusalem was taken not only on the same fast day, that is the same calendar date, but also on the same day; that appears to mean the same day of the week.  Now 27 years is almost exactly 334 lunar months, and 334 lunations require 9,863 days, 5 ¼ hours.  Since 9,863 days is a multiple of seven, every date in the Jewish calendar in 36 B.C. would fall on the same day of the week it did in 63 B.C.  This coincidence would not apply to 37 B.C., however, so the twenty-seven years interval that Josephus gives looks like being genuine. 

Historians have usually accepted the consular dates and rejected the twenty-seven years as an error for two main reasons: firstly, because only by so doing can Herod’s reign be brought to an end as early as 4 B.C., as required by the eclipse, and secondly, because support is supposed to be found for the earlier date in Dio’s Roman History. Actually a superficial reading of Dio’s account of the capture of Jerusalem by Sosius appears to date this event in the consulship of Claudius and Norbanus, 38 B.C., not 37, but a more careful examination reveals that Dio is giving a year-by-year account of the activities of Antony.  He says that after making peace with Antiohus at Samosata on the Euphrates, Antony returned to Italy leaving Sosius in charge of Syria.  He then inserts, as a parenthesis, a thumb-nail sketch of the career of Sosius which ends with his capture of Jerusalem and the downfall of Antigonus.[12]  This is followed, most misleadingly, by the words “This was the course of events in the consulship of Claudius and Noranus” (i.e. 38 B.C.).[13]  But actually this date refers back to the departure of Antony for Italy and the appointment of Sosius as general in charge of Syria. 

This becomes evident from what follows, for Dio goes onto say that Sosius did nothing in 37 B.C. “During the following year (i.e. 37 B.C.)”, he says, “the Romans accomplished nothing worthy of note in Syria.  For Antony spent the entire year in reaching Italy and returning again to the province; and Sosius...spent the time in devising means, not for achieving some success and incurring his enmity, but for pleasing him without engaging  in any activity.”[14]  According to Dio, then, Sosius did nothing in 37 B.C., and since the consular date for 38 B.C. does not apply to Sosius, there is no support at all here for the date given by Josephus, but rather a rebuttal of it.  It follows, then, that Sosius captured Jerusalem in 36 B.C. 

Now the sources from which Josephus drew his historical material were mainly Jewish, and so would not include dates in terms of Roman consuls or Greek Olympiads.  These two dates, namely for the appointment of Herod as king and for the overthrow of Antigonus is Jerusalem, are not give in his earlier work, The Jewish War, although the same events are reported.[15]  It is clear that Josephus added them in the Antiquities, and probably obtained them by conversion from Jewish or Seleucid dates.  But owing to the Roman and Greek years starting in January and July, while the Jewish and Seleucid years began in Nisan or Tishri, he might easily have made an error of one year. Even Schurer, who accepts the consular dates, is obliged to confess that the 184th Olympiad is impossible for the appointment of Herod as king by the Romans.[16]  Apart from this, several other consular dates given by Josephus are impossible to reconcile with one another. 

In fact the dates 40 and 37 B.C. for the accession of Herod are at variance with the chronology of this period as given by Josephus himself.  In a list of high priests and the periods for which they held office, he gives Hyrcanus twenty-four years and Antigonus three years and three months.[17]  Now Hyrcanus was appointed by Pompey in 63 B.C.,[18] whence we deduce that Antigonus began his reign in 39 and was removed by Sosius in 36.  These two terms of office together total twenty-seven years, and so conform to the twenty-seven years’ interval between the two captures of Jerusalem by Pompey and Sosius which historians reject.  Neither can it be argued tha the twenty-four years of Hyrcanus were inclusive of an accession year, for, as we shall see, Josephus did not use that system of reckoning, and furthermore, Antigonus’ term of office is given as three years and three months, which is clearly factual. 

Another entirely independent calculation leads to the same result.  It is agreed that Julius Caesar was murdered in 44 B.C., and that his assassins were defeated by Octavius Caesar and Antony at the battle of Philippi towards the end of 42 B.C.[19]  Now Josephus mentions this battle[20] and mentions that Antony then marched into Asia where he met and fell in love with Cloepatra.[21]  This must have been in 41 B.C., and he goes on to relate how Antony at this time appointed Herod and his brother Phasselus tetrarchs.  It was two years after this, he says,[22] after the Parthians had meanwhile conquered Syria, that they deposed Hyrcanus as high priest, and made Antigonus both king and high priest.  Two years after 41 B.C. is 39 B.C. and it is only then that Herod went to Rome where he interviewed Antony, and got himself appointed king in place of Antigonus.  Once again Josephus does not support his own consular dates. 

Correlation with Roman History 

A date of 39 B.C. for Herod’s appointment as king by the Romans fits into Roman history better than 40 B.C., as the following table of Antony’s movements after the battle of Philippi (42 B.C.) shows: 

41 B.C. Antony went into Asia, met Cleopatra in Cilicia and returned with her to Egypt, leaving Saxa in charge of Syria (Dio, xlviii, 24). 

40 B.C. Parthian invasion of Syria under Libienus and Pacorus; defeat of the Roman general Saxa; Parthian conquest of Syria except Tyre (Dio xlviii, 24-25).  Antony, hearing of Parthian conquests, returned from Egypt to Tyre, but went on to Greece and finally to Italy, where he became involved in hostilities with Caesar’s forces at Brundisium (Dio xlviii, 27).  After the death of Antony’s wife Fulvia, a peace was made in October and Antony married Octavia (Dio xlviii, 28).  If Herod was made king this year, it could only have been after October, and even then the friendly relations between Caesar and Antony, described by Josephus,[23] are at variance with their mutual distrust described by Dio. 

39 B.C.  In Asia the Parthian conquest extended to Palestine; Hyrcanus was deposed and Antigonus made ruler (Dio xlviii, 26, under the account of the previous year, but he states, lxviii, 34, that the foregoing covered two years).  In Italy, following trouble between Caesar and Sextus over Sicily, a peace treaty was signed between Caesar, Antony, and Sextus (Dio xlviii, 36-38). At this time Heord might well have visited  Antony in Rome and been made king.  In fact Appian actually names Herod among several petty kings that were appointed by Antony in this year.[24]  Finally Antony left Italy for Greece where he stayed the winter, sending Ventidius to Asia to deal with the Parthians (Dio xlviii, 39).

38 B.C.  Ventidius, having defeated Labienus (Dio xlviii, 40), was again victorious over the Parthians under Pacorus (Dio, xlix 19-20).  Later Antony arrived from Greece and sent Ventidius home.  He attacked Antiochus at Samosata on the Euphrates, but finally made peace (Dio, xlix, 22).  Finally Antony returned to Italy, leaving Sosius in charge of Syria (Dio xlix, 22). 

37 B.C. Antony spent most of the year in going to Italy, while Sosius was doing nothing in Syria (Dio, xlix, 23).  Exchange of forces in Italy between Antony and Caesar (Dio, xlix, 54), after which Antony returned to Asia (Plutarch, Antony, 45).  There he married Cleopatra and prepared for his conquest of Parthia (Plutarch, Antony, 36-37).   

36 B.C. Antony’s disastrous campaign in Parthia (Dio, xlix, 24-31) 

Meanwhlie Herod’s activities during the years 39 to 36 B.C. as related by Josephus would have been as follows: 

39 B.C. The Parthians under Pacorus and Barzapharnes invaded Judea[25] and some weeks after Pentecost captured Hyrcanus and Phasaelus; meanwhile Herod escaped to Masada where he left his relatives in safety.[26]  About a month after Pentecost Antigonus was made king and high priest by the Parthians.  Herod sought help first in Arabia and then in Egypt, before going to Rome via Pamphylia and Rhodes; there he interviewed Antony and was declared king by the Roman Senate.[27]  Meanwhile Antigonus besieged Masada and bought off Ventidius, the Roman general who had been sent to drive out the Parthians.[28]  Toward the end of the year Herod arrived back from Italy, mustered an army in Galilee, and marched sought to rescue his relatives at Masada, after which he went into winter quarters.[29]  “About the same time Antony was staying at Athens”, says Josephus,[30] agreeing with Dio xlviii, 39 above. 

38 B.C. In the new year Herod conducted a campaign against cave robbers, while Pacorus and the Parthians were defeated by Ventidius.[31]  Herod, unable to make progress against Antogonus, went and joined Antony at the siege of Samosata.[32]  At the end of the year Antony left for Egypt and gave orders to Sosius to assist Herod.[33] 

37 B.C. When Herod came back from Samosata he first mustered an army in Lebanon and marched south to Jericho to revenge his brother Joseph who had been killed there. After that he conducted campaigns against Antigonus in various parts of Judea and Samaria,[34] ultimately confining Antigonus to Jerusalem before the winter came on.[35] 

36 B.C. After a siege lasting five months Jerusalem was taken with the help of the Romans under Sosius, probably on the Day of Atonement, 10 Tishri, Antigonus having reigned three years and three months.[36]  Antigonus was taken captive and sent to Antony as soon as he arrived back from Parthia at the end of the year. 

Confirmation by Sabbatic Years 

Josephus records that the year in which Jerusalem was besieged was sabbatic,[37] and that the resulting food shortage continued even after Herod had become sole ruler. Sabbatic years ran from Tishri to Tishri, so it follows that if Jerusalem was taken in Tishri 36 B.C., the sabbatic year in question ran from Tishri 37 to Tishri 36, and the consequent food shortage would have continued until the harvest of 35.  Alternatively, if Jerusalem was taken in 37 B.C., the sabbatic year was one year earlier. Since sabbatic years occurred in intervals of seven years, a consideration of the dates of other recorded sabbatic years should decide the issue between the two dates. 

An earlier sabbatic year, ending in 150 of the Seleucid ear (160 B.C.)[38] is mentioned in 1 Maccabees in connection with the Maccabean revolt.  Owing to the Second Book of Maccabees giving the date of this as 149 B.C.,[39] it was at one time thought that the Seleucid era was calculated from Nisan 312 B.C. in the First Book of Maccabees but from Nisan 311 in the Second. But W. Kolbe’s opinion[40] that both books use the same era starting from Nisan 311 seems now to have been generally accepted.  J. C. Darcy,[41] however, believes that I Macc. iv. 20 giving 150 B.C. must be an error.  But since one of his main reasons for rejecting this date is that it would be incompatible with a sabbatic year in 37 when Herod captured Jerusalem, his argument is circular and therefore invalid.  A sabbatic year ending in Tishri 162 (150 S.E.) would however be compatible with a sabbatic ending in 36 B.C., the interval being 126 (7 x 18) years, which Josephus himself gives as the duration of the Hasmonean dynasty.[42]   

Another argument against Darcy’s view is that he confesses that he is quite unable to reconcile his sabbatic year in 149 S.E. with that mentioned in connection with the death of the high priest Simon, and the succession of his son, Hyrcanus.  I Macc. XVI. 14 ff, tells us that Simon went down to Jericho in the eleventh month Shebat, in the 177th year of the Seleucid era, and that shortly afterward he was murdered and two of his sons were imprisoned in the fortress of Dok or Dagon.  Josephus adds that when Hyrcanus was besieging the fortress in an attempt to rescue his brothers, “the year one which the Jews used to rest came on,”[43] and so he was obliged to give up the siege.  This statement is peculiar, because there is no reason to think that the Jews ceased from war as soon as a sabbatic year came on.  What Josephus probably meant is that the effects of the sabbatic year came on, namely food shortage.  This would have made itself felt first in the arid mountainous regions near Jericho where the fortress was situated, thus compelling Hyrcanus to withdraw to Jerusalem where he would find stores of food.  The siege of Dagon began in the spring of the year beginning  Nisan 178 S.E/ (134 B.C.) and the food shortage became acute in the summer of that year.  The sabbatic year in question ran, therefore, from Tishri 135 to Tishri 134 B.C. , exactly twenty-eight year after that in the time of the Maccabean revolt, provided that this was in 150 S.E. (162 B.C.) and not 149. 

Against this it has been argued that according to Jewish tradition a sabbatic year ended in A.D. 69, the year before the destruction of Jerusalem.  It is true that the Talmudic Tractate Ta’anith (292, p. 154) seems to imply that both the first and second temples were destroyed in a year following a sabbatic year; but it is only fair to point out that another Tractate, “Arakim (11b, p. 65), states equally clearly that on both occasions that the temple was destroyed at the end of a sabbatic year.  Consequently the Jewish Encyclopedia says that “the exact year of the Shemittah is in dispute and different dates are given.”[44]  Both statements in the Talmud will be found on examination to be founded on superstition, the only solid fact being that on both occasions the temple was destroyed on the same date, 9th Ab: both writers evidently seek to extend this coincidence to the same year in the sabbatic cycle.  The argument that later sabbatic years during the Christian era seem to support A.D. 69 as sabbatic is worthless, because the Jews began to count a new era from the destruction of Jerusalem, reckoning A.D. 70 as year 1, thus making the year 69 appear to be the 7th year of the previous cycle. 

The balance of evidence, therefore, seems to favour sabbatic years in 150 and 178 S.E. (162 and 134 B.C.), and these confirm that a sabbatic year would have fallen in 36 B.C., not 37. 

Herod’s Reign 

Josephus says that Herod reigned thirty-seven years from his appointment by the Roman Senate, or thirty-four years from the overthrow of Antigonus. To calculate the date of his death from these figures, it is necessary to know what system of reckoning Josephus used.  Under the system most commonly used in western Asia, called the “accession-year” system, the regnal years began to count from new year’s day next after the king’s accession, the foregoing fraction being termed his accession year.  According to this system the last regnal year was numerically the same as the length of the reign.  Under the “non-accession-year” system, the accession year counted as the king’s first year, so that all subsequent years counted one year higher, and the length of the reign, again taken as numerically the same as the last year, appeared to be one year longer.[45]  The accession-year is clearly the more practical, since the interval between two events separated by several reigns could be readily calculated by simply adding together the reigns of the intervening kings.  Under the non-accession-year rule, straightforward addition would give an excess of one year for every reign involved.   

This fact provides a simple means of checking which system Josephus used. It is agreed both by Josephus[46] and I Maccabees[47] that Simon became high priest in 170 S.E. (142  B.C.), and a second well-established date is 63 B.C., when Pompey captured Jerusalem and reinstated Hyrcanus as high priest.[48]  The interval seventy-nine years.  During this period there were six priestly rulers whose reigns are given by Josephus as follows: 

Simon               8 years Ant. XIII, vii, 4

Hyrcanus I        31 years Ant. XIII, x, 7; War I, ii, 8

Aristobulus       1 year Ant. XIII, xi, 3; War I, iii, 6

A. Janneus        27 years Ant.XIII, xi, 3; War I, iv, 8

Alexandra         9 years Ant. XIII, xvi, 6; War I, v, 4

Artistobulus      3 ½ years Ant. XIV, vi, 1 

Total                79 ½ years 

If each of these reigns had been reckoned by the non-accession-year system, the total would have exceeded the actual period by six years, and the fact that it does not do so proves that Josephus used the accession-year system.  Following the above period Josephus adds that Hyrcanus II reigned twenty-four years and Antigonus three years and three months,[49] together 17 ½ years, the exact interval between the summer of 63 and the autumn of 36 B.C., when Herod captured Jerusalem. 

In general it would be necessary to know whether Herod’s regnal years began on 1 Nisan or 1 Tishri, but it so happens that his is immaterial for calculating the year of Herod’s death.  We can be fairly certain that both Herod’s capture of Jerusalem as well as his appointment in Rome fell after I Tishri and before Nisan.  Josephus says that Jerusalem was captured “on the solemnity of the fast,” evidently meaning the Day of Atonement, 10 Tishri.  When the Parthians invaded Palestine three years earlier, Herod and the Jews were still in possession of Jerusalem at Pentecost, and Antigonus was not made king until a month after that, for he reigned three years and three months ending on 10 Tishri. This if Herod arrived in Rome before 1 Tishri, he would have less than three months for his flight to Masada with his family, his negotiations with the king of Arabia, his visit to Cleopatra in Egypt and finally his journey to Rome, which was delayed by bad weather that caused a diversion to Rhodes where he was obliged to get a new ship. 

It follows, then, that Herod’s first regnal year, counting from his appointment in Rome, began in 38 B.C. either on I Nisan or I Tishri, and his 37th and last year began in Nisan or Tishri 2 B.C. Alternatively, counting from his capture of Jerusalem on 10 Tishri, 36  B.C., his first regnal year began in 35 B.C., either in Nisan or Tishri, and his 34th and last year in Nisan or Tishri 2 B.C.  In either case his death would have been at the end of January 1 B.C., if he died on 1 Shebat, as the Megilat Ta’amit suggests.  Thus by accepting the revised dates for Herod’s accession, and applying in a straightforward manner the most elementary rules of chronology current in western Asia, we find that Herod died a little over a fortnight after the total eclipse of the moon on the night of 9 January, 1 B.C. 

Herod’s Age 

Josephus informs us that Herod was about 70 years of age when he died.[50]  This provides a further check on the date of his death, for we are told he was 25[51] years of age when his father Antipater made him governor of Galilee.[52]  This occurred shortly after Antipater had been made administrator of Judea by Caesar, as he was on his way from Egypt to Asia Minor in 47 B.C.[53]  From this it follows that Herod would have been 70 in 2 or 1 B.C., but not more than 67 or 68, if he died in the spring of 4 B.C. 

Schurer’s Chronology 

According to current opinion, Herod was declared king by the Romans towards the end of 40 B.C., and died less than 35 ½ years later in the spring of 4 B.C.; alternatively, he reigned as king from his capture of Jerusalem in the autumn of 37 B.C. until the spring of 4  B.C., a period of 32 ½ years.  Since these periods are given by Josephus as 37 and 34 years respectively, it becomes necessary to explain the discrepancy of over 1 ½ years in each case.  In order to do this E. Schurer,[54] whose chronology is generally accepted, makes the following three assumptions:

1.      That Josephus always reckons reigns or periods of time inclusively, that is by the non-accession-year rule.

2.      That Herod’s regnal years began on 1 Nisan.

3.      That Herod died after 1 Nisan in 4 B.C., and that Josephus reckoned the odd day or two of the new years as a full regnal year. 

The first converts fractions of a year into a whole, while the last two make a couple of days count as a year. 

In support of his first assumption he quotes the interval between the capture of Jerusalem by Pompey in 63 B.C. and its capture by Herod and Sosius in 36 B.C. given by Josephus as twenty-seven years.  But as he presupposes his own date, 37, for Herod’s capture, he is able to argue that the period should have been twenty-six years, and therefore Josephus has reckoned inclusively.  The argument is manifestly circular and invalid.  In both of the other examples he cites, eh also assumes his own dates for Herod’s accession.  Against this there is the overwhelming evidence adduced above that Josephus reckoned reigns by the accession-year rule, and that the interval between the two captures of Jerusalem was actually twenty-seven years. 

To support his view that Herod’s regnal years began in Nisan and that the first few days of this month would have been counted as a full year, Schurer quotes Rosh Hashanah 1, 1 in the Talmud.  It is true that either by accession-year or non-accession-year reckoning the first few days of a new year would count as a full year, but unfortunately the statement that the new year for kings was 1 Nisan is supported only by a fallacious argument purporting to show that Solomon’s regnal years began in Nisan.  It has been clearly demonstrated by E. R. Thiele[55] that not only did Solomon’s regnal years begin on 1 Tishri, but so also did those of all the kings of Judah down to Zedekiah.  The cuneiform record of the Babylonia capture of Jerusalem in 597 B.C. confirms this in the case of Zedekiah.  It is there stated that the city was taken and a new king (Zedekiah) appointed in the month Adar[56]  If the new year for kings had been on 1 Nisan, as it was for Babylonian kings, then Zedekiah’s first year would have begun on 1 Nisan, 597, and would have been identical with Nebuchadrezzar’s 8th year. But this is impossible because Zedekiah’s 10th and 11th years are equated with Nebuchadrezzar’s 18th and 19th (Jer. xxxii.1 and 2 Kings xxv.2 and 8).  It follows that Zedekiah’s first year did not begin until Tishri 597, so that the second halves of his 10th and 11th years would overlaps the 18th and 19th of Nebuchadrezzar. 

If then the reigns of Solomon and the kings of Judah are to be regarded as precedent for Herod’s regnal years, we must reject Schurer’s theory that they began in Nisan, and even if they did begin in Nisan there is no evidence that Herod died in that month.  We find, then, that Schurer’s chronology allows Herod a reign of only thirty-six or thirty-three years, not thirty-seven or thirty four as Josephus requires. 

Synchronism with Roman History 

At several points it is possible to synchronize dated events in Herod’s reign with known dates in Roman history.  But before this can be done, it is necessary to know whether Herod’s regnal years, as given by Josephus, have been calculated from his appointment in Rome, or from his overthrow of Antigonus three years later.  According to Schurer’s chronology Herod’s first year began when he overthrew Antigonus, supposedly in the middle of 37 B.C., and on that basis agreement can be achieved between the dates given by Josephus and dates in Roman history.  There is reason to believe, however, that Josephus did not regard this as the official beginning of Herod’s reign, but counted from his appointment in Rome, for throughout the chapter recording Herod’s activities after his return from Rome, and before he captured Jerusalem, he repeatedly called Herod “the king.”[57] 

This view is supported by Herod’s coins.  It has been pointed out by B. Kansel that coins always express conditions de jure and not de facto, so that in this case they would be dated from Herod’s legal appointment in Rome, and not from his taking possession of Jerusalem.[58]  But the peculiar fact is that of all the coins of Herod’s reign, only those of the 3rd year are dated, and furthermore, appear to commemorate some event of great importance.  This could only have been the capture of Jerusalem, for in the third year after that event nothing of any consequence is known to have happened. The conclusion is that Herod, in minting these coins, wished to emphasize that he had already been king three years when he captured Jerusalem, and that his regnal years must be reckoned accordingly. 

Assuming then that Josephus regarded Herod’s appointment in Rome late in 39 B.C. as the time of his accession, and that he applied the accession-year rule with the new year on 1 Tishri, Herod’s first regnal year ran from Tishri, 38 to Elul, 37 B.C.  According to Schurer’s non-accession-year system, on the other hand, counting Nisan years from Herod’s capture of Jerusalem, Herod’s first year ran from Nisan, 37 to Adar, 36 B.C.  It will be seen, therefore, that the six months, Nisan to Elul, 37 B.C. would be in Hero’s first year by either system of reckoning.  Similarly throughout Herod’s reign events falling in the summer half year would be dated in the same regnal year by either system, and consequently synchronization with Roman history would be unaffected. 

The three events principally concerned are (a) an earthquake before the battle of Actium (3 September, 31 B.C.) in Herod’s 7th year,[59]  (b) the expedition of Aelius Gallus in 24 B.C.[60] for which Herod provided 500 men in his 14th year,[61]  and (c) Herod’s announcement in his 18th year of his intention to rebuild the temple;[62] this can be dated from the visit of Augustus to Syria in 20 B.C.[63]  All these remain valid under the revised chronology. As regards the earthquake, Josephus says this occurred “in the early spring.”[64]  If this means before I Nisan, then Schurer’s chronology fails, for it would have been in Herod’s 6th year by his reckoning. 

Regarding the rebuilding of the temple, however, there is one statement that cannot be reconciled with Schurer’s chronology.  In The Jewish War[65] Josephus gives the date of Herod’s start on this work as the 15th, not the 18th year of his reign. It was pointed out some time ago in this journal that this contradiction could be explained if Josephus had used two sources, one dating events from Herod’s appointment in Rome, the other from his capture of Jerusalem.[66]  The revised chronology allows this explanation, but on the basis of Schurer’s chronology the 15th year must be summarily rejected as an error.[67] 

Herod’s Successors     

One of the chief reasons for supposing that Herod died in 4 B.C. is that his sons who succeeded him appear to have begun their reigns in that year.  Thus Archelaus, ruler of Judea and Samaria, was banished in A.D. 6/7 after a reign of ten years;[68] Antipas, tetrarch of Galilee and Perea, who died in A.D. 39 or 40, reigned, according to coin evidence, forty-three years,[69] while Philip, tetrarch of Iturea, died in the 20th year of Tiberius, A.D. 33/34, after a reign of thirty-seven years.[70]  This kind of indirect evidence, however, can be misleading, and has often been a cause of error in the case of the kings of Israel and Judah.  On several occasions, when a king appointed his son as co-regent, the son’s reign overlapped the father’s by several years.  It appears that Herod did the same at least in the case of Antipater, for in testifying to Varus about him he said, “to whom I have in a manner yielded up my royal authority while I am yet alive.”[71] Likewise Antipater in his reply said, “I was king already…you proclaimed me king in your lifetime.”[72]  Antipater was deposed for the murder of his uncle Pheroras, tetrarch of Galilee, and was later executed.  Most of the kingdom was then given to Archelaus, but it is not clear how long this was before Herod died. 

After Herod’s death Archelaus declared, with false humility, that he was not properly king until his succession had been confirmed by the Roman emperor. He went to Rome where his enemies brought self-contradictory charges against him, arguing at one moment that Herod had not appointed him king until he was already demented on his death bed,[73] at another that “he had long exercised royal authority”.[74]  The latter may well be nearer the truth, but without any precise chronological information it is difficult to give a time-table.  The following outline would account for the stated reigns of Archelaus and Antipas: 

4 B.C. Pheroras, tetrarch of Galilee, was murdered by Antipater and succeeded by Antipas, whose reign of forty-three years ended in A.D. 39/40.  Later Antipater’s plot ws discovered and Archelaus nominated as king in his place, his reign of ten years ending in A.D. 6/7. 

3  B.C. Antipater, residing in Rome, was not informed for seven months that he had been charged with murder,[75] and when at last he did return to Palestine his journey was not at all hurried.  It would be late in the year before he was brought to trial before Varus, governor of Syria. Even then Herod hesitated to pass sentence, contenting himself with only sending a report to Rome.[76] 

2 B.C. As a result of intercepting Antipater’s correspondence further conspiracies came to light and, after more procrastination, Herod sent more ambassadors to Rome[77] who returned only a few days before his death in January 1  B.C.  At some time during this year Herod tried to disinherit Archelaus, but his last will shows that this was only temporary. 

We still have to account for Philip who, according to Josephus, died in the 20th year of Tiberius after he had reigned as tetrarch for thirty-seven years.[78]  The context  shows, however, that this is almost certainly an error.  It seems that a figure has been dropped, and that the text should probably read the 22nd year of Tiberius.  The chapter in question begins by telling us that Vitellius, a “man that had been consul”, became president of Syria and visited Jerusalem at the time of the Passover.  Vitellius was consul in A.D. 34, so he could not have arrived in Jerusalem until the spring of 35, before going on to Antioch. During that year we are told that he engineered a war to drive the Parthians out of Armenia.  In the following year he met Artabanus, king of Parthia, on the Euphrates and negotiated a peace.[79]  It was about this time, we are told, that Philip died. Now if he reigned thirty-seven years ending in A.D. 36, his reign must have begun in 1 B.C., the year that we have already found to be that of Herod’s death, following which Philip was made tetrarch by the emperor Augustus in Rome. 

Now the 20th year of Tiberius ended in August 34, while Vitellius was still consul in Rome, so this date cannot be right.  The 22nd year, however, would be correct, and in this connexion F. Riess[80] quotes the Franciscan Mokenbhr[81] as saying that he had see early copies of Josephus, one a Parisian copy dated 1517 and another Venetian copy dated 1481, in which the text reads “the 22nd year of Tiberius”.  If this could be verified, it would not only clear up a difficult passage in Josephus, but make if difficult to argue from this text that Herod died in 4  B.C. 


Josephus provides a great deal of chronological material regarding the reign of Herod, but owing partly to errors and partly to ambiguities, it is possible to correlate the data with two quite different chronologies.  Some of the evidences that have been regarded as conclusively in favour of 4 B.C. as the date of Herod’s death, such as the lunar eclipse, are seen to be entirely neutral.  There are, of course, weaknesses on both sides, but it is submitted that Schurer’s dates for the accession of Herod in 40 or 37  B.C. cannot be upheld, and since it is impossible to accept parts of both chronologies, the whole of his chronology must be rejected.  If, as seems likely, Herod’s accession was noe year later, this would lead to the conclusion that he died in January, 1 B.C.  W.E. Filmer


[1] Ant. XVII, vi. 4 (167) and ix. 3.

[2] Finegan, Handbook of Biblical Chronology, § 458, table 140.

[3] E. Shurer, Geschichre des Schresches Volech (1901) Vol. 1, p. 416, n. 167.

[4] See Schegg, (1882)

[5] F.K. Ginzel (1809) p. 146 n.os 960 and 962.

[6] Ginzel no. 963.

[7] S. Zeitlin, Megillot Ta’anit, p. 109.

[8] Ant. XVII, vi, 5.

[9] Ant. XIV. XLV.

[10] Appian, Civil Wars, V, 75 (chapt. vii).

[11] Ant. XIV. XVI, 4..

[12] Dio, xlix,  22.

[13] Dio, xlix. 23, 3.

[14] Dio, xlix. 23. 1-2.

[15] War, I, xiv, 4.

[16] E. Schurer, History of the Jewish People, Vol. I,  p. 393, n. 3.

[17] Ant. XX, x.

[18] Ant. XIV, iv, 4.

[19] C.A.H. 5, p. 24

[20] Ant. XIV, xii, 2.

[21] Ant. XIV, xiii, 8.

[22] Ant. XIV, xiii, 3.

[23] Ant, XIV, xiv, 4-5.

[24] Appian, Civil Wars v. 75 (chapt. viii).

[25] Ant. XIV, xii, 3; War I, xiii, 1.

[26] Ant. XIV, xiii, 4-9.

[27] Ant. XIV, xiv, 1-5.

[28] Ant. XIV, xiv, 6.

[29] Ant. XIV, xv, 1-4.

[30] Ant. XIV, xv, 5.

[31]Ant. XIV, xv, 5-7.

[32] Ant. XIV, xv, 8.

[33] Ant. XIV, xv, 9.

[34] Ant. XIV, xv, 10-12.

[35] Ant. XIV, xv, 13 and 14.

[36] Ant. XIV, xvi and xx. 2.

[37] Ant. XIV, xvi, 2.

[38] I Macc. IV, 40 and 53.

[39] II Macc. XIII, 1.

[40] Hermes (Berlin), vol lxii (1927), pp. 225 ff.

[41] Commentary on I Maccabees, p. 51.

[42] Ant. XIV, xvi, 4.

[43] Ant. XIII, viii, 1.

[44] Jewish Encyclopedia X, pp. 607 ff.

[45] E. R. Thiele, Mysterious Numbers of the Hebrew Kings, pp. 14f.

[46] Ant. XIII, vi, 7.

[47] I Macc. xlii. 41-42.

[48] C. A. H. ix, pp. 38a f.

[49] Ant. XX, x

[50] Ant. XVII, vi, 2; Wars, I, xxxiii, 2.

[51] The Greek text reads fifteen years, but his must be an error, for otherwise Herod could never have reached the age of seventy.

[52] Ant. XIV, ix, 2.

[53] C. A. H. ix, p. 678.

[54] Schurer, History of the Jewish People, vol. I, p. 465, n. 165.

[55] E. R. Thiele, Mysterious Numbers of the Hebrew Kings,, pp. 31 f.

[56] D. J. Weiseman, Chronicles of Chaldean Kings, pp. 33 and 73.

[57] Ant. XIV, xv.

[58] Jewish Quarterly Review, vol. xlii (1951/2), pp. 261 ff.

[59] Ant. XV, v, 2.

[60] Dio, xii,  29, 3.

[61] Ant. XV, ix, 1 and 3.

[62] Dio, liv, 7, 4-6.

[63] War I,  xix, 3.

[64] War I, xxi, 1.

[65] War I, xxi, 1.

[66] J.T.S. vol. xxxvi (1935), pp. 25f.

[67] J. Finegan, Handbook of Biblical Chronology, § 432.

[68] Dio lv,  27, 6; Ant. Xvii, xiii, 2.

[69] E. Schurer, History of the Jewish People, vol. ii, pp 36f, n. 45.

[70] Ant. XVI, iv, 6.

[71] War, I, xxxii, 2.

[72] War, I, xxxii, 3.

[73] War II, ii, 5; Ant. XVII, ix, 5.

[74] War, II, ii, 5.

[75] Ant. XVII, iv, 3; War I, xxxi, 2.

[76] Ant. XVII, v, 7.

[77] Ant. XVII, v, 7-8; War I, xxxii, 6-7.

[78] Ant. XIII, iv, 6.

[79] C.A.H. x, pp. 718f; Ant. XVIII, iv, 3-5.

[80] F. Reiss, Das Giburtafer Christi, p. 54.

[81] Dissert. Critica de annis quibus Christus erit natus (Monastarii, 1798), p. 135.



Adoration of the Shepherds


All rights reserved.

Top of page