Find the average of two or more percentages the right way. Enter each percentage below — and, when the groups behind them are different sizes, turn on sample sizes to compute a proper weighted average. You cannot simply average percentages from unequal groups, and this calculator handles the weighting for you.
The average percentage calculator finds the mean of two or more percentages — but, crucially, it does it correctly. Averaging percentages sounds like it should be as simple as adding them up and dividing by how many there are, and sometimes it is. But when each percentage comes from a group of a different size, that simple approach gives the wrong answer. This calculator lets you switch on sample sizes so that larger groups carry proportionally more weight, producing the true overall percentage. Enter your percentages, add the group sizes if you have them, and the tool returns the accurate average along with the full working.
Below you will find a clear explanation of when a simple average is fine and when you need a weighted one, the exact formulas, worked examples you can reproduce, the most common mistakes people make, and answers to frequently asked questions. By the end you will know not just what the answer is, but exactly why averaging percentages needs a little more care than averaging ordinary numbers.
Sometimes — but only when the groups behind the percentages are the same size. Suppose two classes of exactly 30 students each take the same test, and one class scores an average of 80% while the other scores 40%. Because the classes are equal in size, the overall average really is (80% + 40%) ÷ 2 = 60%. The simple average works here because each percentage represents the same number of people.
Now change the situation slightly. Suppose 80% is the score of a group of four people and 40% is the score of a single person. A naive average still gives 60%, but that is wrong. Four people scored 80% and only one scored 40%, so the four high scorers should count four times as much as the single low scorer. The correct average is (80×4 + 40×1) ÷ 5 = 360 ÷ 5 = 72%, not 60%. This is the heart of the matter: percentages can only be averaged directly when they represent groups of equal size. Whenever the group sizes differ, you must use a weighted average.
When every percentage represents the same number of items, the average is the ordinary (arithmetic) mean:
average = (a₁ + a₂ + a₃ + ... + aₙ) ÷ n
Here a₁, a₂, and so on are your percentages and n is how many of them there are. You add all the percentages together and divide by the count. For example, the simple average of 20%, 40%, and 90% is (20 + 40 + 90) ÷ 3 = 150 ÷ 3 = 50%. Use this only when you are confident the underlying groups are equal in size, or when the percentages are not tied to groups at all — for instance, averaging three separate discount rates you are considering.
When the groups differ in size, each percentage must be weighted by its sample size — the number of items or people in that group. The formula is:
weighted average = (a₁·w₁ + a₂·w₂ + a₃·w₃ + ... + aₙ·wₙ) ÷ (w₁ + w₂ + w₃ + ... + wₙ)
Here each a is a percentage and each w is the size of the group it came from. You multiply every percentage by its sample size, add up all those products, and then divide by the total of all the sample sizes. The larger a group, the more it pulls the average toward its own percentage. If all the weights happen to be equal, this formula collapses back to the simple average — the simple mean is just a weighted mean where every group is the same size.
Imagine you surveyed three age groups about whether they like pancakes and got these results: 64% of 300 young people, 42% of 450 middle-aged people, and 36% of 250 older people said yes. What percentage of everyone surveyed likes pancakes? Because the groups are different sizes, you need a weighted average:
(64×300 + 42×450 + 36×250) ÷ (300 + 450 + 250)
= (19,200 + 18,900 + 9,000) ÷ 1,000
= 47,100 ÷ 1,000 = 47.1%
So 47.1% of everyone surveyed likes pancakes. These are the calculator's default values — press Calculate to see the full breakdown, then enter your own figures. Notice that a naive average of 64%, 42%, and 36% would give 47.33%, which is close but not exactly right, because it ignores that the middle group was the largest.
The gap between a simple and a weighted average grows when the group sizes are lopsided. Suppose a store reports a 90% customer-satisfaction rate from a survey of 900 shoppers on weekdays, and a 50% rate from just 100 shoppers on a chaotic sale weekend. A simple average suggests (90 + 50) ÷ 2 = 70%. But the weighted average tells the real story:
(90×900 + 50×100) ÷ (900 + 100) = (81,000 + 5,000) ÷ 1,000 = 86,000 ÷ 1,000 = 86%
The true overall satisfaction is 86%, not 70%, because the vast majority of shoppers were surveyed on weekdays. Reporting 70% here would badly misrepresent the business. This is exactly the kind of error that weighting prevents.
Choosing correctly is usually easy once you ask one question: does each percentage stand for the same number of things? If yes, a simple average is fine. If no, use a weighted average. A few concrete cases help:
When in doubt, weighting by sample size is the safer choice, and if the groups turn out to be equal it will give the same answer as the simple average anyway.
A percentage throws away information: it tells you a proportion but hides the size of the group it came from. "80%" could describe four people out of five or eight hundred out of a thousand. When you average percentages without their group sizes, you implicitly treat every group as equally important, which quietly gives tiny groups the same influence as huge ones. In real data this can flip conclusions entirely — a small, unrepresentative sample with an extreme percentage can drag a naive average far away from the truth. Weighting restores the missing information by letting each group count in proportion to how many observations it actually contains. That is why statisticians, pollsters, and analysts almost always weight percentages by sample size when combining them.
There is a neat way to check a weighted average percentage: it should equal the total of the "yes" counts divided by the total number of observations. In the pancake example, the number who like pancakes is 64% of 300 (192 people) plus 42% of 450 (189 people) plus 36% of 250 (90 people), which is 471 out of 1,000 — exactly 47.1%. This shows that a weighted average of percentages is really just a shortcut for going back to the underlying counts, pooling them, and computing one overall percentage. If you happen to have the raw counts, you can always compute the combined percentage directly; the weighted-average formula simply lets you do it when all you have are the percentages and the group sizes.
Averaging percentages correctly comes up constantly. Teachers combine class averages into a grade-level average and must weight by class size. Marketers blend conversion rates across campaigns of different traffic volumes. HR teams roll up department satisfaction scores where departments vary in headcount. Retailers combine regional sales-growth percentages weighted by each region's revenue. Researchers pool response rates from studies with different sample sizes in a meta-analysis. Public-health officials combine vaccination or infection rates across populations of very different sizes. In every one of these cases, using a simple average would understate the influence of the largest groups and could lead to a materially wrong headline figure — which is exactly why a weighted average percentage is the standard tool.
It helps to remember what a percentage really is: a proportion multiplied by 100. Because percentages are just scaled proportions, everything that is true about averaging proportions is true about averaging percentages. The mean of a set of proportions, weighted by the number of observations behind each, gives the overall proportion — and multiplying by 100 turns it back into a percentage. This is why the weighted-average formula is the same one used for any weighted mean, whether the quantities are prices, test scores, speeds, or rates. Once you see percentages as ordinary numbers that happen to be measured "per hundred," the rules for combining them stop feeling like a special case and become just another application of the weighted mean.
If the groups are equal in size, add the percentages and divide by how many there are. If the groups differ in size, use a weighted average: multiply each percentage by its group size, add the products, and divide by the total of the group sizes.
Because that treats every group as equally important. If the groups are different sizes, a small group counts as much as a large one, which distorts the result. You need to weight each percentage by its sample size.
It is an average where each percentage is multiplied by the size of the group it represents before being combined, so larger groups have proportionally more influence on the final figure.
When every percentage represents the same number of items, or when the percentages are standalone rates not tied to any group size. In those cases the simple and weighted averages are identical.
Yes. Use the "Add another percentage" button to enter as many percentages as you like. The formula extends naturally to any number of entries.
No. Just type the number — enter 47 for 47%. The calculator treats every entry as a percentage automatically.
This Average Percentage Calculator is provided for general educational and informational purposes. It computes the standard simple and weighted averages of the percentages you enter. For sequential percentage changes (such as compounding growth rates), a different method applies, as noted above.