Random Number Table

Random number table with 5-digit blocks, a seed for repeatable results, and steps to draw a random sample or assign groups. Print it or copy it. Free.

Chart your own dataPaste raw numbers into the Histogram Maker to bin them and draw the bars.
0114137570059708741551769639636728604890323443002361
0245333880294089438937668163085858677367115903139791
0392259602553946507244141621263381279279996530228257
0493267287735195650359129125063592534162110535200171
0595104319971363376568390540350582799360585812039052
0691101762555523728254939685299716357067600390324377
0766461361427423541743247729510401253530127563599200
0869528324113310592103103075275089413200817858424841
0901033055648503183351578167413890419452215292439069
1098907300345248934084402467003253488570602553318551
1163642780092559421741086753404727788865330705371422
1217598929167476411329970255851267388533823281168711
1338319996190884202738490964290531612288320928303695
1412017741450427765353851673668053480904336155257626
1513765578296701304392899979144636016745202439576239
1690204107368116780874573607765114019668688906067271
1758970956433413204544188444849960582501720020126596
1819803027981691236556679773532433953207313747870421
1996892566815460836621081769062844014110294053488022
2009263690432572988661780212347687588038748319434104

Pseudo-random digits (mulberry32, seed 1955), not the RAND table and not for cryptography or lotteries.

Number your population 1 to N, then read the table above. The reader takes as many digits at a time as N has, runs straight on across block gaps and into the next row, and skips 0, anything above N, and numbers already chosen. You can also click any digit in the table to start there.

Reading 2 digits at a time (N = 80)
StepRow, digitReadResult
Press “Read next” to take the first number.

Chosen so far (0 of 8)

None yet

Free Random Number Table Charts

Every chart below is drawn from the same code as the tool above. View, download and print them free, no sign-up.

20 rows

Random Number Table: 5-Digit Blocks, Seed 1955

The default table: 20 rows, ten blocks of five digits each, generated from seed 1955. The same seed always gives the same table.

1 / 4

Tables

How to read

Checks

Randomization

To look up the cutoff for the digit check, use the chi-square table. Have your own numbers to chart? Paste them into the histogram maker. Not sure which table fits your test? The statistics tables guide sorts z, t, chi-square and F tables by test.

History source: RAND Corporation, A Million Random Digits with 100,000 Normal Deviates (online edition, including the description of how the digits were produced).

What a random number table is

  • A random number table is a page of digits in which every digit from 0 to 9 is equally likely at every position, independent of the digits around it. Before computers were common it was the standard way to pick a random sample or assign people to treatment groups: you number your population, enter the table at an arbitrary point, and let the digits decide.
  • The digits above come from a seeded pseudo-random generator (mulberry32), not from physical noise. Type a seed and you get the same table every time, on any device, which is what makes a sample or an allocation list reproducible and auditable. The default seed is 1955. Change it with the New seed button, or share the link, which carries the seed and the table size.

How to read the table for a random sample

  • Number the members of your population 1 to N. Read as many digits at a time as N has: two for N = 80, three for N = 250. Start at a row and digit you did not choose by looking (or press Random start), read left to right across the row, and carry on into the next row. The gaps between blocks are only there to help your eye, so ignore them.
  • Skip 0, skip anything larger than N, and skip any number you already have. Stop when you have n different members. If the table ends first, the reader stops and says so instead of wrapping around, because wrapping would repeat the same digits.
  • Worked example: choose 8 of 80 members starting at row 1, digit 1. Reading two digits at a time gives 14, 13, 75, 70, 05, 97 (skip), 08, 74, 15. The sample is members 14, 13, 75, 70, 5, 8, 74, 15.

Random assignment to groups

  • Simple random assignment gives each participant an independent random group, like flipping a coin per person. It is unbiased, but with small samples the groups can end up unequal.
  • Blocked (permuted block) randomization fixes this. Participants are cut into blocks of 4 or 6, and each block contains every group equally often in a random order, so group sizes never drift far apart: with two groups they differ by at most half a block at any point. With seed 1955 and groups A and B in blocks of 4, the first block is A B B A. One caveat: if staff know the block length they can predict the last assignments in a block, so trials that must stay masked often vary the block length.

Are the digits really uniform?

  • The Digit check tab counts how often each digit appears and runs a chi-square goodness-of-fit test against equal frequencies, with 9 degrees of freedom. For the default table the counts give χ² = 7.92 and p = 0.5422, so there is no sign of unevenness. A single low p-value is not a defect: about 1 table in 20 falls below 0.05 by chance alone. If you want a different table, change the seed rather than searching for a seed that looks good, which defeats the purpose.

The RAND table and why this one is different

  • The best-known printed table is A Million Random Digits with 100,000 Normal Deviates, published by the RAND Corporation in 1955. RAND generated its digits in 1947 with an electronic roulette wheel built from a random pulse source, then rerandomized them by adding digits modulo 10 because the raw output showed slight departures from randomness.
  • This page does not reproduce that table. Its digits are pseudo-random: computed from a seed by an algorithm, so anyone with the seed can regenerate them exactly. That is useful for teaching and for documenting a classroom or lab allocation, but it is not suitable for cryptography, security tokens, or regulated lotteries and draws, which need a certified or cryptographically secure source.

Frequently Asked Questions