Skip to content
Statculator

Statistical table

Pearson r Critical Value Table

Pearson r critical value table for df = n − 2 from 1 to 1000, one- and two-tailed, with an exact lookup, p-value check, worked examples and a free PDF.

Updated

Jump to table Download PDF

Quick answer

An r critical value table lists the smallest absolute Pearson correlation that is statistically significant for a given sample size and significance level. Rows are degrees of freedom, df = n − 2, and columns are α levels. If the absolute value of your sample r is at least the tabled value, reject H₀: ρ = 0. For n = 20 (df = 18) at two-tailed α = 0.05, the critical value is 0.4438.

Find the critical r for your sample

The table rows use df = n − 2.

Shows the exact p-value and the decision.

Critical r, n = 20 (df = 18), two-tailed α = 0.05

±0.4438

A correlation is significant if |r| ≥ 0.4438.

Pearson r table: critical values of the correlation coefficientTop row: one-tailed α. α₂: two-tailed α. Rounded to 4 decimals.
One-tailed α0.10.050.0250.010.0050.0005
df = n − 2α₂ = 0.20α₂ = 0.10α₂ = 0.05α₂ = 0.02α₂ = 0.01α₂ = 0.001
10.95110.98770.99690.99950.99991.0000
20.80000.90000.95000.98000.99000.9990
30.68700.80540.87830.93430.95870.9911
40.60840.72930.81140.88220.91720.9741
50.55090.66940.75450.83290.87450.9509
60.50670.62150.70670.78870.83430.9249
70.47160.58220.66640.74980.79770.8983
80.44280.54940.63190.71550.76460.8721
90.41870.52140.60210.68510.73480.8470
100.39810.49730.57600.65810.70790.8233
110.38020.47620.55290.63390.68350.8010
120.36460.45750.53240.61200.66140.7800
130.35070.44090.51400.59230.64110.7604
140.33830.42590.49730.57420.62260.7419
150.32710.41240.48210.55770.60550.7247
160.31700.40000.46830.54250.58970.7084
170.30770.38870.45550.52850.57510.6932
180.29920.37830.44380.51550.56140.6788
190.29140.36870.43290.50340.54870.6652
200.28410.35980.42270.49210.53680.6524
210.27740.35150.41320.48150.52560.6402
220.27110.34380.40440.47160.51510.6287
230.26530.33650.39610.46220.50520.6178
240.25980.32970.38820.45340.49580.6074
250.25460.32330.38090.44510.48690.5974
260.24970.31720.37390.43720.47850.5880
270.24510.31150.36730.42970.47050.5790
280.24070.30610.36100.42260.46290.5703
290.23660.30090.35500.41580.45560.5620
300.23270.29600.34940.40930.44870.5541
350.21560.27460.32460.38100.41820.5189
400.20180.25730.30440.35780.39320.4896
450.19030.24290.28760.33840.37210.4647
500.18060.23060.27320.32180.35420.4432
600.16500.21080.25000.29480.32480.4079
700.15280.19540.23190.27370.30170.3798
800.14300.18290.21720.25650.28300.3568
900.13480.17260.20500.24220.26730.3375
1000.12790.16380.19460.23010.25400.3211
1500.10450.13390.15930.18860.20830.2643
2000.09050.11610.13810.16360.18090.2298
3000.07400.09480.11290.13380.14800.1884
5000.05730.07350.08750.10380.11490.1464
10000.04050.05200.06190.07350.08130.1038

What is an r critical value table?

Pearson’s correlation coefficient r measures the strength of a linear relationship in a sample. Even when two variables are unrelated in the population (ρ = 0), sample correlations scatter around zero by chance, and in small samples they can be surprisingly large: with 5 pairs, r has to reach 0.8783 before it is significant at the 5% level. The critical value is the cutoff beyond which a sample r would be unlikely to arise by chance alone at the chosen significance level.

The same table is published as a Pearson r table, a correlation coefficient table or correlation coefficient chart, a table of Pearson correlation critical values, critical values for Pearson correlation and a critical value r table. All of them hold the same numbers: the critical value of r for each df and α. It is not a table of correlation coefficients between your variables (that is a correlation matrix).

Rows give df = n − 2 from 1 to 30, then 35, 40, 45, 50, 60, 70, 80, 90, 100, 150, 200, 300, 500 and 1000. Columns give two-tailed α = 0.20, 0.10, 0.05, 0.02, 0.01 and 0.001, and the row of small numbers above them gives the matching one-tailed α, which is half of each. Values are rounded to four decimals. The test assumes a roughly linear relationship and approximately bivariate normal data. For any other n or α, or to get a p-value for your own r, use the lookup above the table.

How to read the r critical value table

  1. Count the number of pairs n and compute df = n − 2.
  2. Decide between a two-tailed test (H₁: ρ ≠ 0) and a one-tailed test (H₁: ρ > 0 or ρ < 0). For a one-tailed test at α, read the column whose one-tailed α (top row) matches; it is the two-tailed 2α column.
  3. Find the df row and the α column, and read the critical value r_c.
  4. Compare the absolute value of your sample correlation with r_c. Reject H₀ if |r| ≥ r_c; for a one-tailed test, the sign of r must also match H₁.
  5. If your df is not listed, use the next smaller df, which gives a slightly larger critical value and so a conservative test, or enter n in the exact lookup.

Examples

Two-tailed test

Problem. For n = 20 pairs, r = 0.45. Is the correlation significant at α = 0.05, two-tailed?

  1. 01df = 20 − 2 = 18.
  2. 02Row 18, two-tailed 0.05 column: r_c = 0.4438.
  3. 03Compare: |0.45| = 0.45 ≥ 0.4438.

Answer. Yes, narrowly: reject H₀ (the exact two-tailed p-value is about 0.046).

Checking two significance levels

Problem. A study of n = 12 pairs finds r = −0.62. Is the correlation significant at α = 0.05? At α = 0.01?

  1. 01df = 12 − 2 = 10.
  2. 02Two-tailed α = 0.05: r_c = 0.5760. Since |−0.62| = 0.62 ≥ 0.5760, the correlation is significant.
  3. 03Two-tailed α = 0.01: r_c = 0.7079. Since 0.62 < 0.7079, it is not significant at this level.

Answer. The negative correlation is significant at the 5% level but not at the 1% level (exact two-tailed p ≈ 0.032), so report 0.01 < p < 0.05.

One-tailed test

Problem. Before collecting data, a researcher predicted a positive correlation. With n = 30 pairs they find r = 0.35. Test at one-tailed α = 0.05.

  1. 01df = 30 − 2 = 28.
  2. 02One-tailed α = 0.05 is the two-tailed 0.10 column: r_c = 0.3061.
  3. 03r is positive, as predicted, and 0.35 ≥ 0.3061.

Answer. Reject H₀: there is evidence of a positive correlation (one-tailed p ≈ 0.029). A two-tailed test at 0.05 would not reject, because 0.35 < 0.3610.

Critical values of r for common sample sizes

Most correlation tests use α = 0.05 or 0.01, two-tailed, so these are the values you will look up most often. This short Pearson r table is indexed by the number of pairs n rather than df; the full table above has every df from 1 to 30 and selected values up to 1000.

Pairs ndf = n − 2Two-tailed α = 0.05Two-tailed α = 0.01One-tailed α = 0.05
530.87830.95870.8054
1080.63190.76460.5494
15130.51400.64110.4409
20180.44380.56140.3783
25230.39610.50520.3365
30280.36100.46290.3061
40380.31200.40260.2638
50480.27870.36100.2353
100980.19660.25650.1654

Where the critical values come from

The table is derived from the t-test for a correlation. Under H₀: ρ = 0, the statistic t = r√(n − 2) / √(1 − r²) follows a t distribution with n − 2 degrees of freedom. Solving that equation for r at the critical t value t_c gives the tabled critical correlation:

  • r_c = t_c / √(t_c² + df)
  • Example: df = 18 and two-tailed α = 0.05 give t_c = 2.1009 (from the t-table), so r_c = 2.1009 / √(2.1009² + 18) = 0.4438.

One-tailed vs two-tailed r table

A two-tailed test asks whether the correlation differs from zero in either direction; a one-tailed test asks about one direction chosen before seeing the data. Because the sampling distribution of r under H₀ is symmetric, the one-tailed critical value at α equals the two-tailed value at 2α. Use the one-tailed reading only if a correlation in the opposite direction would be treated the same as no correlation.

Two-tailed αOne-tailed αr_c at df = 10r_c at df = 28
0.200.100.39810.2407
0.100.050.49730.3061
0.050.0250.57600.3610
0.020.010.65810.4226
0.010.0050.70790.4629
0.0010.00050.82330.5703

Finding a p-value range from the r table

Read across your df row and find the two critical values your |r| falls between: the p-value lies between their α levels. With n = 12 (df = 10) and r = −0.62, the row reads 0.5760 at two-tailed α = 0.05 and 0.6581 at 0.02, so 0.02 < p < 0.05; the exact value is 0.032. If |r| is above the 0.001 column, report p < 0.001. Type your r into the lookup above, or use the Pearson r to P-Value Calculator, for the exact p-value. For what a p-value does and does not tell you, see What Is a P-Value? Meaning, Calculation and Common Mistakes.

Significant is not the same as strong

With large samples, weak correlations become significant. At df = 100 the two-tailed 0.05 critical value is 0.1946, and at df = 1000 it is only 0.0619, a correlation that explains less than 0.4% of the variance (r² ≈ 0.0038). From about 385 pairs, even r = 0.1 is significant at the 5% level. Report r, r² and a confidence interval, and judge strength in context.

To test a correlation with full working, use the Pearson’s Correlation Test Calculator; to compute r from raw data, use the Correlation Coefficient Calculator; and for an exact critical value at any n, use the Pearson r Critical Value Calculator.

Print or download the r table

Use Download PDF at the top of the page for a printable copy: the Pearson r critical values are Table 5 (page 10) of the free Statculator statistical tables PDF, which also holds the z, t, chi-square and F tables. The PDF rounds to three decimals; Print / save as PDF prints this page at four decimals with no scroll box.

Every value is computed from the exact Student t inverse CDF, so it can differ from older printed tables in the last digit. You can check the t values behind any entry against the NIST/SEMATECH e-Handbook table of critical values of Student’s t, and the test itself is described in Penn State STAT 462, section 2.6.

Frequently asked questions

What is the critical value of r?

It is the smallest absolute sample correlation that is statistically significant for your sample size and α. It depends on df = n − 2: at two-tailed α = 0.05 it is 0.6319 for 10 pairs, 0.4438 for 20 pairs and 0.3610 for 30 pairs. A larger sample needs a smaller r to be significant.

How do I use an r critical value table?

Compute df = n − 2, find that row, and read the value in the column for your α. If the absolute value of your sample correlation is at least that critical value, the correlation is statistically significant. For n = 12 at two-tailed α = 0.05, the critical value is 0.5760.

Why is df = n − 2 for a correlation?

Testing a Pearson correlation is equivalent to testing the slope of a simple linear regression, which estimates two parameters: the intercept and the slope. Each estimated parameter uses up one degree of freedom, leaving n − 2.

What is the critical value of r for n = 30?

With df = 28, the two-tailed critical values are 0.3610 at α = 0.05 and 0.4629 at α = 0.01. For a one-tailed test at α = 0.05, the critical value is 0.3061.

How do I get the critical value of r in R or Excel?

Neither has a built-in function, so convert from t. In R, tc <- qt(1 - 0.05/2, 18) gives 2.1009, and tc / sqrt(tc^2 + 18) gives 0.4438. In Excel, put =T.INV.2T(0.05, 18) in A1, then =A1/SQRT(A1^2+18). Replace 18 with your n − 2 and 0.05 with your two-tailed α.

Does a significant correlation mean a strong relationship?

No. Significance means the population correlation is unlikely to be exactly zero. With about 385 or more pairs, even r = 0.1 is significant at the 5% level, so judge strength from the size of r and r² and from a confidence interval.

Can I use this table for Spearman’s rank correlation?

Only as a rough guide for larger samples. Spearman’s rank correlation has its own exact critical values for small n, which differ from Pearson’s. The Spearman Correlation Calculator tests the significance of a rank correlation directly.

Cite this page

Statculator. (2026, November 14). Pearson r Critical Value Table. Retrieved November 14, 2026, from https://statculator.com/tables/r-critical-value-table/

Values are computed from the exact distribution functions used by Statculator’s calculators (see methodology).