Statistical table
F-Table (F Distribution Table)
F distribution table of critical values for α = 0.10, 0.05, 0.025 and 0.01 by numerator and denominator df, with an exact lookup, ANOVA examples and a free PDF.
Updated
Quick answer
An F table gives upper-tail critical values of the F distribution for a chosen α, with the numerator degrees of freedom df₁ across the columns and the denominator degrees of freedom df₂ down the rows. Pick the table for your α, find the cell for (df₁, df₂), and reject H₀ if your F statistic is larger. For α = 0.05, df₁ = 3 and df₂ = 20, the critical value is 3.098.
Find any F critical value
Shows the right-tail p-value.
Upper F critical value, α = 0.05
3.0984
F(0.05; 3, 20), area α to the right.
What is an F-table?
The F distribution describes the ratio of two independent variance estimates, each divided by its degrees of freedom. It appears whenever a test compares variances: one-way ANOVA (between-group against within-group mean squares), the overall F-test in regression, and the F-test for equality of two variances. F is never negative and is skewed to the right, so its critical values come from the upper tail.
The same table goes by several names: F distribution table, F-test table, F statistic table, F critical value table, F chart or "ftable", and, after the two statisticians behind the distribution, the Fisher–Snedecor distribution table. The number you read from it is called the tabulated F value or F crit, and you compare it with the F value calculated from your data.
Because the distribution has two degree-of-freedom parameters, each significance level needs its own two-way grid. This page has four F tables, for α = 0.10, 0.05, 0.025 and 0.01. Columns give the numerator df₁ (1–10, 12, 15, 20, 24, 30, 40, 60, 120 and ∞) and rows give the denominator df₂ (1–30, 40, 60, 120 and ∞). Each cell is the value F_c with P(F ≥ F_c) = α. For any other df or α, for left-tailed and two-tailed values, or for an exact p-value, use the lookup above the table.
How to read the F-table
- Work out both degrees of freedom. In one-way ANOVA, df₁ = k − 1 (number of groups minus one) and df₂ = N − k (total observations minus number of groups).
- Select the table for your significance level with the α buttons. For a two-sided test of two variances at α = 0.05, use the α = 0.025 table for the upper critical value.
- Find the column for df₁, the numerator degrees of freedom.
- Find the row for df₂, the denominator degrees of freedom. Do not swap them: F(3, 20) and F(20, 3) have very different critical values.
- Read the cell, and reject H₀ if the calculated F is greater than this tabulated F value. If a df is missing, use the exact lookup or the rule in the FAQ below.
Examples
One-way ANOVA
Problem. An ANOVA compares k = 4 groups with N = 24 observations in total and gives F = 3.45. Is the result significant at α = 0.05?
- 01df₁ = k − 1 = 3 and df₂ = N − k = 20.
- 02In the α = 0.05 table, column 3 and row 20 give 3.098.
- 03Compare: 3.45 > 3.098.
Answer. Reject H₀ at α = 0.05: at least one group mean differs (the exact p-value is about 0.036).
Overall F-test in regression
Problem. A multiple regression with m = 4 predictors is fitted to n = 35 observations. What F is needed for significance at α = 0.01?
- 01df₁ = m = 4.
- 02df₂ = n − m − 1 = 35 − 4 − 1 = 30.
- 03In the α = 0.01 table, column 4 and row 30 give 4.018.
Answer. The regression is significant at the 1% level only if F > 4.018.
Two-sided test of two variances
Problem. Samples of n₁ = 11 and n₂ = 13 are used to test H₀: σ₁² = σ₂² against H₁: σ₁² ≠ σ₂² at α = 0.05, with F = s₁² / s₂². Find both critical values.
- 01df₁ = 10 and df₂ = 12. Each tail gets α/2 = 0.025, so use the α = 0.025 table.
- 02Upper critical value: column 10, row 12 gives 3.374.
- 03Lower critical value: swap the degrees of freedom, read column 12, row 10 (3.621), and take the reciprocal: 1 / 3.621 = 0.276.
Answer. Reject H₀ if F < 0.276 or F > 3.374. Choosing "Two-tailed" in the lookup gives the same pair.
Which α table should you use?
ANOVA and regression F-tests are always right-tailed, so use the table whose α matches your significance level. The two-variance F-test is different when the alternative is two-sided: a ratio far below 1 is as much evidence as one far above 1, so α is split across both tails and you read the α/2 table.
| Test | Tail | Table at a 5% significance level | df₁, df₂ |
|---|---|---|---|
| One-way ANOVA | Right | α = 0.05 | k − 1, N − k |
| Overall F-test in regression | Right | α = 0.05 | m, n − m − 1 (m predictors) |
| Two variances, H₁: σ₁² > σ₂² | Right | α = 0.05 | n₁ − 1, n₂ − 1 |
| Two variances, H₁: σ₁² ≠ σ₂² | Both | α = 0.025 for the upper value, reciprocal rule for the lower | n₁ − 1, n₂ − 1 |
F table at α = 0.05: most-used critical values
Most F-tests are run at the 5% level with a handful of groups or predictors, so these are the values you will look up most often. Rows are the denominator df₂ and columns the numerator df₁. For example, 2 and 15 degrees of freedom give 3.682, and 2 and 16 give 3.634. The full α = 0.05 grid, with df₂ 1–30, 40, 60, 120 and ∞, is in the table above.
| df₂ ↓ / df₁ → | 1 | 2 | 3 | 4 | 5 | 6 | 8 | 10 |
|---|---|---|---|---|---|---|---|---|
| 5 | 6.608 | 5.786 | 5.409 | 5.192 | 5.050 | 4.950 | 4.818 | 4.735 |
| 10 | 4.965 | 4.103 | 3.708 | 3.478 | 3.326 | 3.217 | 3.072 | 2.978 |
| 12 | 4.747 | 3.885 | 3.490 | 3.259 | 3.106 | 2.996 | 2.849 | 2.753 |
| 15 | 4.543 | 3.682 | 3.287 | 3.056 | 2.901 | 2.790 | 2.641 | 2.544 |
| 16 | 4.494 | 3.634 | 3.239 | 3.007 | 2.852 | 2.741 | 2.591 | 2.494 |
| 20 | 4.351 | 3.493 | 3.098 | 2.866 | 2.711 | 2.599 | 2.447 | 2.348 |
| 24 | 4.260 | 3.403 | 3.009 | 2.776 | 2.621 | 2.508 | 2.355 | 2.255 |
| 30 | 4.171 | 3.316 | 2.922 | 2.690 | 2.534 | 2.421 | 2.266 | 2.165 |
| 60 | 4.001 | 3.150 | 2.758 | 2.525 | 2.368 | 2.254 | 2.097 | 1.993 |
| 120 | 3.920 | 3.072 | 2.680 | 2.447 | 2.290 | 2.175 | 2.016 | 1.910 |
Finding a p-value with the F table
One F table cannot give an exact p-value, but the four α tables together can bracket it. Look up the same (df₁, df₂) cell in each table and see which two critical values your F statistic falls between: the p-value lies between those two α levels.
Example: F = 3.45 with df₁ = 3 and df₂ = 20. The cell reads 2.380 at α = 0.10, 3.098 at 0.05, 3.859 at 0.025 and 4.938 at 0.01. Because 3.098 < 3.45 < 3.859, the p-value is between 0.025 and 0.05; the exact value is 0.036. If F is above the α = 0.01 value, report p < 0.01; if it is below the α = 0.10 value, report p > 0.10. Type your statistic into the lookup above, or use the P-Value from F Calculator, for the exact p-value.
Calculated F vs tabulated F (F crit)
The calculated F comes from your data, for example MS_between / MS_within in an ANOVA table. The tabulated F, often labelled F crit, comes from this page and depends only on α, df₁ and df₂. The decision rule is the same for every right-tailed F-test: if calculated F > tabulated F, reject H₀.
Software prints the same tabulated value. Excel’s ANOVA output has an "F crit" column, and =F.INV.RT(0.05, 3, 20) returns 3.098; in R, qf(0.95, 3, 20) gives the same number. These match the α = 0.05 table at column 3, row 20.
Left-tailed and two-tailed F critical values
F tables print only upper-tail values. To get a lower-tail critical value, use the reciprocal identity: the value with area α below it in F(df₁, df₂) equals 1 divided by the upper-tail α value of F(df₂, df₁). Swap the degrees of freedom, look up the upper value, and take its reciprocal. For a two-tailed test, do this with α/2 for the lower value and read the α/2 table for the upper value. The Tail setting in the lookup does both steps for you.
A common shortcut avoids lower critical values altogether. Put the larger sample variance in the numerator, so that F ≥ 1, use that sample’s df as df₁, and compare F with the upper α/2 critical value.
Relationship to the t-table
When df₁ = 1, the F critical value is the square of the two-tailed t critical value with the same denominator df. In the α = 0.05 table, F(1, 10) = 4.965, which is 2.228² up to rounding, and 2.228 is the two-tailed 5% value for df = 10 in the t-table. This is why a two-group ANOVA and a pooled t-test always reach the same conclusion.
For an exact critical value at any df or α, use the lookup on this page or the F Critical Value Calculator. To run the full analysis, use the ANOVA Calculator or the F-Test Calculator. If you are not sure which ANOVA design you have, and so which degrees of freedom apply, see One-Way vs Two-Way ANOVA: Differences and When to Use Each.
Print or download the F table
Use Download PDF at the top of the page for a printable copy: the four F tables are Tables 4a–4d (α = 0.10, 0.05, 0.025 and 0.01, pages 6–9) of the free Statculator statistical tables PDF, which also holds the z, t, chi-square and Pearson r tables. Print / save as PDF prints this page with all four F tables and no scroll box.
Every value is computed from the exact F inverse CDF and rounded to three decimals, so it can differ from older printed tables in the last digit. You can check any entry against the NIST/SEMATECH e-Handbook table of upper critical values of the F distribution, which uses the same layout; its F distribution page gives the formulas.
Frequently asked questions
Which degrees of freedom go across the top of an F table?
What is the F table value at 0.05?
What is the F critical value for α = 0.05 with 2 and 12 degrees of freedom?
Why is there a separate F table for each α?
How do you find a p-value from an F table?
What if my degrees of freedom are not in the F table?
Can an F statistic be less than 1?
Cite this page
Statculator. (2026, November 14). F-Table (F Distribution Table). Retrieved November 14, 2026, from https://statculator.com/tables/f-table/
Skip the table: exact calculators
- F Critical Value CalculatorF critical values for any df₁, df₂ and α.
- ANOVA CalculatorOne-way ANOVA for 2–10 groups with the full ANOVA table, Tukey HSD and effect size.
- F-Test CalculatorTest whether two population variances are equal using F = s₁² / s₂².
- P-Value from F Calculatorp-value for an F statistic with numerator and denominator df.
Values are computed from the exact distribution functions used by Statculator’s calculators (see methodology).