Bar graph vs pie chart: which to use
By Jude Wallis · Published
A bar graph and a pie chart both display one categorical variable. A bar graph shows each category's frequency as the length of a separated bar, which makes comparing categories easy. A pie chart shows each category as a slice of a whole circle, which emphasizes each part's share of the total.
AP Statistics: Unit 1 (topics 1.4). Bar charts and pie charts are the graphical representations for one categorical variable in topic 1.4 of Unit 1 in the Fall 2026 AP Statistics course. The CED notes a pie chart's slice areas represent relative frequencies and together equal 100% of the whole.
Bar graph vs pie chart: the short answer
Both graphs display the same kind of data: one categorical variable, a variable that sorts each observation into a group such as favorite sport or eye color. They differ in what they make easy to see. A bar graph turns each category's count into the length of a bar, so your eye compares lengths and ranks the categories quickly. A pie chart turns each category's share into a slice of one circle, so your eye reads each category as a fraction of the whole.
Because people judge length more accurately than angle or area, the bar graph is usually the better choice, especially when you want to compare categories or when there are more than a few of them. The pie chart earns its place when the point is that the categories are parts of a single whole that add up to 100%.
What a bar graph shows
A bar graph, also called a bar chart, draws one bar per category, with the bar's height (or length, if drawn sideways) equal to the frequency (count) or relative frequency (proportion) in that category. The bars are separated by gaps, which signal that the categories are distinct groups rather than points on a number line.
For an unordered (nominal) categorical variable, the order carries no statistical meaning, so you may arrange the bars by size or alphabetically and rearranging them never changes what the graph says. When the variable is ordinal, such as small, medium, and large, keep the bars in that natural sequence, because the ordering is part of what the variable records. Bar graphs also extend cleanly to comparing two or more groups side by side, which a single pie chart cannot do. The histogram versus bar graph guide covers why a bar graph's gaps matter and how it differs from the quantitative histogram it resembles.
What a pie chart shows
A pie chart displays the frequencies or relative frequencies of a categorical variable as slices of a single circle. Each slice stands for one category, and the area of the slice, as a fraction of the whole circle, equals that category's relative frequency. All the slices together fill the circle, so their shares sum to 1, or 100%.
That part-to-whole framing is the pie chart's one real strength: it makes the message "these categories divide up a fixed total" visible at a glance. The cost is that comparing two similar slices, or reading an exact value, is hard, because people estimate angles and areas poorly. A pie chart also needs the categories to be non-overlapping and to cover every case, or the slices will not add to a full circle.
The differences side by side
| Feature | Bar graph | Pie chart |
|---|---|---|
| Data type | One categorical variable | One categorical variable |
| Encodes value as | Length of a bar | Angle or area of a slice |
| Best at | Comparing and ranking categories | Showing parts of one whole |
| Number of categories | Handles many | Best with only a few |
| Compare across groups | Yes, side by side | Not in one chart |
| Reading exact values | Easier | Harder |
Both show the same categorical counts. The split is length versus slice, which is why the bar graph wins for comparison and the pie chart wins for part-to-whole.
When to use which
Reach for a bar graph by default. Use it whenever you want to compare category sizes, rank them, read values with any precision, handle more than about four categories, or place two groups next to each other. Most of the categorical questions you meet fit this description.
Use a pie chart only when the single most important message is that the categories are shares of one fixed total and there are just a few of them. Even then, label each slice with its percentage, because the angles alone are hard to read. If you find yourself squinting to compare two slices, that is the signal to redraw the data as a bar graph.
The classic mix-up and how to avoid it
The frequent mistakes are drawing a pie chart when the categories do not form a whole, and using a pie chart to compare categories it cannot separate clearly. A pie chart of "languages a student speaks" breaks down if students can speak more than one, because the overlapping categories will not sum to 100%. And a pie with eight near-equal slices hides the very ranking a reader wants.
Avoid both with two checks. First, confirm the categories are non-overlapping and cover every observation before you consider a pie, so the slices can add to one whole. Second, ask what the reader should take away: if it is a comparison or a ranking, use a bar graph, and if it is a few parts of a single total, a labeled pie is defensible. Neither graph is for quantitative data, which belongs on a histogram or dotplot instead, as the quantitative versus categorical guide explains.
Turning counts into bars and pie slices
Fifty students report how they get to school: walk 20, bus 15, car 10, bike 5. Show how the same counts become a bar graph and a pie chart, computing the relative frequencies and the pie slice angles.
Confirm the total: students, and the four categories do not overlap, so a pie chart is valid.
For a bar graph, the heights are just the counts: walk 20, bus 15, car 10, bike 5, drawn as four separated bars in any order.
For relative frequencies, divide each count by 50: walk , bus , car , bike .
Check the proportions sum to one: .
For pie slice angles, multiply each relative frequency by 360 degrees: walk , bus , car , bike .
Check the angles sum to a full circle: degrees.
Compare readability: from the bars you can see at once that walking is twice as common as car travel ( versus ); from the pie you can see walking fills 40% of the circle, a bit less than half.
The bar graph plots the raw counts (20, 15, 10, 5) as bar heights; the pie chart plots the relative frequencies (0.40, 0.30, 0.20, 0.10) as slice angles of 144, 108, 72, and 36 degrees, which sum to 360. Both show the same data, but the bars make the ranking and the 2-to-1 walk-to-car gap easier to read than the slices do.
Frequently asked questions
When is a pie chart better than a bar graph?
A pie chart is better only when the main point is that a few non-overlapping categories divide one fixed total, and you want that part-to-whole split seen at a glance. For comparing categories, ranking them, or reading exact values, a bar graph is clearer.
Do the slices of a pie chart always add to 100%?
They should, because each slice's area is that category's share of the whole circle, and the shares of all categories cover the total. If your categories overlap or leave cases out, the slices will not add to 100%, which is a sign a pie chart is the wrong graph.
Can a bar graph and a pie chart show the same data?
Yes. Both display the frequency or relative frequency of one categorical variable, so the same counts can be drawn either way. The choice is about which is easier to read for your message, not about which data each can hold.