Margin of error calculator
Enter a poll's sample size to see its margin of error at 95% confidence.
Examples
| Sample | Margin of error (50% result) |
|---|---|
| 400 | ±4.9 |
| 600 | ±4.0 |
| 800 | ±3.5 |
| 1,000 | ±3.1 |
| 1,500 | ±2.5 |
| 2,000 | ±2.2 |
| 5,000 | ±1.4 |
Notice that quadrupling the sample only halves the margin, which is why most national polls interview 800 to 1,500 people.
Questions and answers
What is a poll's margin of error?
How far a poll's result could be from the true value just by chance, because it surveys a sample instead of everyone. At the usual 95% confidence level, a poll of 1,000 people has a margin of about 3.1 points.
How is it calculated?
For a simple random sample: 1.96 × √(p × (1 − p) / n), where n is the sample size and p the proportion (50% gives the largest, most cautious margin).
Does it apply to the lead between two candidates?
No. The margin on the gap between two candidates is roughly twice the margin on each number, because when one goes up the other usually goes down.
What the margin of error doesn't cover
It only measures random sampling error. It doesn't include other errors, such as some groups being less likely to respond, a badly worded question or the difficulty of knowing who will vote, which is why polls sometimes miss by more than their margin. Weighted online polls usually report a somewhat larger "effective" margin than this formula gives.
Margin of error in the glossary · Which pollsters were most accurate in 2026 · Our methodology