Average Percentage Calculator

What Is an Average Percentage?

Averaging percentages sounds like simple arithmetic, but it hides a trap that catches out students, analysts, and journalists alike. If three test scores are 80%, 90%, and 70%, adding them up and dividing by three gives 80%, and that is correct because each test presumably had the same number of questions. But if you are averaging percentages that come from different sample sizes, such as approval ratings from surveys of different populations, that same simple approach can produce a badly wrong figure. The Indeed guide to weighted averages works through exactly this problem and shows why the sample size behind each percentage matters as much as the percentage itself.

This calculator offers three ways to average percentages correctly, so you are not stuck guessing which formula applies to your numbers, and it flags a warning whenever your sample sizes look too different for a simple average to be trustworthy.

Three Calculation Modes

Most average percentage tools carry out only one type of average. This one covers all three common situations you are likely to run into.

Simple average -- use this when every percentage came from the same base or sample size, such as several test scores each out of 100 marks. Add the percentages and divide by how many there are.

Weighted average -- use this when your percentages come from different sample sizes. Enter each percentage alongside its sample size, and the calculator multiplies each percentage by its weight before dividing by the total weight, in line with the method the weighted average percentage method guide sets out.

From raw numbers -- the most reliable option when you have access to the original figures. Enter the part and whole for each group (such as 8 correct out of 10, and 200 correct out of 1,000), and the calculator sums the parts and wholes separately before working out a single overall percentage, sidestepping any weighting approximation entirely.

Why Simple Averaging Goes Wrong With Different Sample Sizes

Given that this is the single most common percentage mistake reported across finance, research, and school settings, it is worth working through a concrete example. The GeeksforGeeks worked examples on averaging percentages use a similar case: imagine a small pilot survey of 10 people found 80% approval, and a much larger follow-up survey of 1,000 people found only 20% approval. A simple average of 80% and 20% gives 50%, which looks like a reasonable midpoint. But the true combined approval rate, calculated from the raw counts (8 approvals out of 10, plus 200 approvals out of 1,000), comes to (8 + 200) divided by (10 + 1,000) times 100, which is 20.59%, nowhere near the naive 50% figure. The small survey's 80% carried the same weight as the large survey's 20% in the simple average, even though it represented a hundred times fewer people.

That said, this does not mean simple averages are always wrong. If every group you are averaging genuinely has the same or a very similar sample size, a simple average and a weighted average will produce nearly identical results, and the extra step is not necessary. The problem only shows up when the underlying bases diverge significantly, which is precisely when this calculator's weighted mode flags a warning, letting you carry out the correction before the figure goes any further.

Real-World Applications by Field

The Investopedia explanation of weighted averages in finance covers the portfolio blending case in the table above in more depth, where allocation size plays exactly the same role that sample size plays in survey research. Our percentage calculator handles the underlying single-percentage arithmetic if you need to work out raw counts before averaging them here.

Choosing the Right Method for Your Data

If you only have the percentages themselves and no access to the original sample sizes, a simple average is the only option available, but it should be reported with the caveat that it assumes equal or similar bases. If you know the sample sizes but not the raw counts, the weighted average mode is the correct choice. If you have the original numerators and denominators, use the raw numbers mode, since it removes any approximation and calculates the true combined percentage directly from first principles. The Calculator Academy guide to averaging percentages walks through this same decision tree with several additional worked examples.

On top of that, it is worth checking your result makes intuitive sense: an averaged percentage should generally fall between your smallest and largest input percentages. If it does not, look into whether a weight or sample size was entered incorrectly, since that is the most common source of an unexpected result. If the underlying figures are themselves percentage changes rather than static percentages, our percentage change calculator is the better starting point before averaging.

Accuracy and Verifying Your Result

This calculator runs on JavaScript floating-point arithmetic, accurate to roughly fifteen significant figures, well beyond the precision most averaging tasks in business or education actually need, in line with the standards the BIPM Guide to the Expression of Uncertainty in Measurement recommends for reporting combined figures. The step-by-step output breaks each mode's calculation into its component sums, so you can narrow down exactly how the final figure was reached and verify it against a spreadsheet if needed.

I find the most reliable habit when averaging percentages is to default to the raw numbers mode whenever the original counts are available, rather than reaching for a percentage-only shortcut. In my experience, this single habit prevents the vast majority of averaging errors I have come across in client reports, since it removes the weighting step as a potential source of mistakes entirely. With that in mind, if you are ever unsure which mode applies, working back to the original parts and wholes is almost always worth the extra minute it takes to figure out.