After the census the total number of Levites does not tally. How can we reconcile the difference?

Numbers 3:21 (KJV)

21 Of Gershon was the family of the Libnites, and the family of the Shimites: these are the families of the Gershonites. 22 Those that were numbered of them, according to the number of all the males, from a month old and upward, even those that were numbered of them were seven thousand and five hundred.

Numbers 3:27 (KJV)

27 And of Kohath was the family of the Amramites, and the family of the Izeharites, and the family of the Hebronites, and the family of the Uzzielites: these are the families of the Kohathites. 28 In the number of all the males, from a month old and upward, were eight thousand and six hundred, keeping the charge of the sanctuary.

Numbers 3:33 (KJV)

33 Of Merari was the family of the Mahlites, and the family of the Mushites: these are the families of Merari. 34 And those that were numbered of them, according to the number of all the males, from a month old and upward, were six thousand and two hundred.

Numbers 3:39 (KJV)

39 All that were numbered of the Levites, which Moses and Aaron numbered at the commandment of the LORD, throughout their families, all the males from a month old and upward, were twenty and two thousand.

The total number of Levites does not tally with the sub totals given of the Gershonites, Kohathites & Merarites. How can we reconcile the above figures?

  • All numbers within the passage are given with precisely two orders of magnitude.
    – Lucian
    Sep 8, 2021 at 5:58

5 Answers 5


This is interesting. The reconciliation for me was with the Kohathites in Numbers 3:28. That was the one verse in which I noted a difference. On this link http://biblehub.com/numbers/3-28.htm the KJV and every other Bible shows the same count of 8,600 Kohathites. However, the NABRE isn’t included among those Bibles; it shows only 8,300 Kohathites.


The NABRE presents a reconciliation (i.e. 7,500 Gershonites + 8,300 Kohathites + 6,200 Merarites = 22,000 total Levites).

In Bible Gateway, ERV is another that has only 8,300[a] Kohathites, but it also has this footnote.

  1. Numbers 3:28 8300 Some copies of the ancient Greek version have “8300.” The Hebrew copies have “8600.” See Num. 3:22, 28, 34, 39.

The total number of the Levite male population according to verse 39, was 22,000 from one month old and upward. The total number of the firstborn males among Israel 22,273. There is apparently, some discrepancy with the number of Levites which is probably the result of some textual error in one or more of the numbers representing the three separate family lines. 2:39 has the total number at 22,000 which is correct but, 7,500 + 8,600 + 6,200 is not 22,000, it is 22,300, In which case the Levites would have outnumbered the first-born of Israel.

Verse 46 says the excess of the firstborn of Israel was 273. Each person over and above the Levites was to pay a five shekels ransom. Verse 50 says the total ransom collected was 1,365 Shekels; 1,365 divided by 5 is 273 so, verse 39 is correct. The error then, is somewhere in the manuscript record of one or perhaps all the three Levite family lines. Every first-born male of Israel was to be represented by a Levite. Since the firstborn of Israel outnumbered the Levite males, a ransom of five shekels per each excess person was required to be paid to the Levites by way of compensation because the Lord claimed all the first-born of Israel. Rather than exercising his claim of all the firstborn, the Lord chose the Levites to serve in their place. Those who were more than the number of Levites were required to redeem themselves with a payment of five shekels.


The total number of Levites given in Numbers 3:22;28;34(22,300), as deduced by adding together the numbers of Kohathites, Merarites, and Gershonites, represented the total number of Levites at that time. The number of Levites given in verse thirty-nine(22,000) represented the total number of Levites that were eligible for purchase by way of the number of firstborn sons which God already owned, having 'purchased' them on the night of Passover by way of preserving them alive(Numbers 3:13, 8:16-19). Since there were twenty-two thousand two hundred and seventy-three firstborn non-Levitical Israelites(v. 43), and only twenty-two thousand Levites to be purchased, the change therefore consisted of two hundred and seventy-three non-Levitical persons, and these were subsequently 'released' by way of their individual donations to the sanctuary(vs. 45-51). Given the above however, it is necessary to conclude that the three hundred extra Levitical persons were themselves already firstborn, and were thus not needing to be seen as being purchased by God, since they, having been firstborn, would have already been 'purchased' on Passover night, according to verse thirteen.


Like oldhermit, I also think that this is a manuscript/scribe copying problem.



it is quite possible that one letter may have been substituted by the scribe for another.


Others consider the difference due to an error in the Hebrew text.


It may, therefore, have arisen from an error in transcription (the alteration of a single letter would suffice);

Keil and Delitzsch

We must rather assume that there is a copyist's error in the number of one of the Levitical families;


The Hebrew root אֶלֶף (elef) means 1,000 as a cardinal number, but multiple scholars have acknowledged it is also used to mean a large group or clan. As noted by Richard Freund:

Linguists have studied this word and decided that the word elef may have only meant in its original, historical usage, a 'legion', a 'clan', or a 'group'...the book of Numbers, chapter 1, verse 16 states: 'these were the ones chosen from the congregation, the leaders of their ancestral tribes, the heads of the clans (elfei) of Israel.' It is clear that the Priestly writer here was using the word elef as a parallel for the word 'tribes', and most translators agree that it probably means 'clans'. (see here)

Some consider it heresy to merely mention the possibility that the numbers are rounded, but heresy is not a topic for this site. The numbers here (and in numerous other OT passages) are rounded to either the nearest hundred or the nearest thousand. A perfect census has never been conducted in modern times--perhaps it is unfair to set a standard for an ancient census-worker that cannot be achieved today.

A far more likely procedure would be to have captains of 100 who reported to captains of 1000. The count then conducted would not be of individuals but of leaders, and the number of leaders (whether they choose to do 10s, 100s, 1000s, etc.) would then be multiplied by the nominal size of that leader's group.

Interesting asides:

  • A Roman "century" (commanded by a centurion) was officially a group of 100 soldiers...but seldom actually had a full 100. It was a group like a modern squad or company with a flexible number of men
  • We still do this today: millipede means "one thousand legs"...they do not in fact have one thousand legs.
  • As noted in other responses, there may in fact be a textual variant here. From the Pulpit commentary:

It may, therefore, have arisen from an error in transcription (the alteration of a single letter would suffice)


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.