Chart Types: A Complete Guide
GUIDEA plain-English tour of the main chart types — what each one shows, when to reach for it, and how to read it.
The main types of charts are bar charts (compare categories), line charts (change over time), pie and donut charts (parts of a whole), area charts (cumulative totals over time), scatter plots (relationships between two numbers) and histograms (distributions). Pick the chart that matches your question, not the one that looks best.
A chart turns numbers into a picture so patterns that are invisible in a table — a trend, a gap, a dominant share — become obvious at a glance. But a chart only works when its shape matches the question you are answering. A pie chart answers "what share of the whole?"; a line chart answers "how did this change over time?"; a scatter plot answers "are these two things related?". Pick the wrong shape and even correct data misleads.
This guide walks through the chart types you will use most often. For each one you get a quick visual, the question it answers, when to use it, and the mistakes to avoid. When you are ready to build one, every type links to a free maker where you can enter your own data and export a PNG or SVG.
How to choose the right chart
Start from the relationship you want to show, not the chart you find prettiest. Four questions cover most cases:
- Parts of a whole? Use a pie or donut chart — but only with a few categories.
- Comparing separate categories? Use a bar chart. It is the most accurate everyday chart for comparison.
- A value changing over time? Use a line chart (or an area chart to stress the cumulative total).
- A relationship between two variables? Use a scatter plot.
When two chart types both seem to fit, choose the one that lets the reader judge values by length or position (bar, line, scatter) rather than by angle or area (pie). People read length far more accurately than they read slices.
Which chart should I use? Quick-pick table
Find the question you are trying to answer in the left column. The middle column is the safest default; the right column lists alternatives and when they beat the default.
| Your question | Best default | Alternatives |
|---|---|---|
| How do separate categories compare? | Bar chart | Horizontal bars for long labels or many categories |
| How do categories compare across two groupings (e.g. region by year)? | Grouped bar chart | Stacked bar if the totals matter more than each part |
| What share of the whole is each part? | Pie chart (2–5 slices) | Donut chart to show the total in the centre; bar chart for 6+ parts |
| How did one value change over time? | Line chart | Area chart to stress volume; column chart for a few periods |
| How did the mix of parts change over time? | 100% stacked bar or stacked area | Small multiples of line charts when each part's trend matters |
| Are two numeric variables related? | Scatter plot | Add a trend line; use colour for a third group |
| How is one variable distributed? | Histogram | Box plot to compare distributions between groups |
| How do a few items score across many attributes? | Radar chart (3–8 axes) | Grouped bar chart for precise comparison |
| When does each task start and end? | Gantt chart | A simple timeline for milestones only |
| Readers need to look up exact values | Table | Table plus a small chart for the headline pattern (see table vs chart) |
Worked example: a box office records 1,840 confirmed and 160 cancelled tickets for one event. The question is "what share was cancelled?" — a part of one whole — so a two-slice pie or donut showing 92% confirmed and 8% cancelled answers it instantly. If the question becomes "is the cancellation rate rising month by month?", the data is now a value over time, and a line chart of the monthly cancellation percentage is the right choice. Same subject, different question, different chart.
Pie & donut charts
The question it answers: what share of the total does each category take? A pie chart divides a single circle so that each slice's angle is proportional to its value, and all slices add up to 100%.
When to use it: when you have one whole split into a small number of parts — say a budget across 3–5 categories, or a survey with a handful of answers — and the takeaway is "this slice dominates" rather than precise comparison.
Too many slices (anything past ~6 becomes unreadable); slices that do not represent parts of a single whole; and 3-D pies, which distort the very angles you are asking readers to compare. If you need exact comparison, use a bar chart instead.
Read the full pie chart guide, or jump straight into the pie chart maker.
Bar & column charts
The question it answers: how do separate categories compare? Each bar's length encodes a value, and because the eye judges length precisely, bar charts are the most reliable everyday chart for comparison.
When to use it: ranking categories (top products, votes per option), comparing groups, or showing a value over a small number of discrete periods. Use vertical bars (columns) for time or short labels, and horizontal bars when labels are long or there are many categories.
Starting the value axis above zero — this exaggerates differences and is the most common way bar charts mislead. Always begin the numeric axis at zero.
Read the full bar chart guide, or open the bar chart maker.
Line charts
The question it answers: how did a value change over time (or another continuous range)? Connecting points with a line emphasises the trend and the rate of change between them.
When to use it: time series — revenue by month, temperature by day, a metric tracked over years. Lines also let you compare several series on the same axes to see which grows fastest.
Connecting categories that have no order (use a bar chart for those), and cramming so many lines onto one chart that none can be followed. Three or four series is usually the practical limit.
Read the full line chart guide, or open the line chart maker.
Area charts
The question it answers: how did a total change over time, with emphasis on the cumulative volume? An area chart is a line chart with the region under the line shaded, which stresses magnitude rather than just the path.
When to use it: a single series where the "amount" matters as much as the trend, or stacked areas to show how a total splits into parts over time. For pure trend comparison, a plain line is cleaner.
The line chart maker includes an area mode — switch the style from Line to Area.
Scatter plots
The question it answers: is there a relationship between two numeric variables? Each point is plotted by its x and y value, so clusters, trends, and outliers become visible.
When to use it: exploring correlation (does height relate to weight?), spotting outliers, or showing the distribution of many data points. A rising cloud suggests a positive relationship; a shapeless cloud suggests none.
Read the full scatter plot guide.
Reading correlation as causation. A scatter plot can show two things move together; it cannot prove one causes the other.
Other useful chart types
Beyond the core five, a few specialised charts answer narrower questions:
- Stacked bar chart — compares categories and shows the parts within each bar.
- Grouped (clustered) bar chart — places related bars side by side to compare sub-categories.
- Histogram — looks like a bar chart but shows the distribution of one continuous variable across ranges (bins).
- Area chart — a line chart with the area beneath filled to stress cumulative magnitude.
- Donut chart — a pie chart with a hollow centre that can hold a total.
- Gantt chart — a horizontal-bar timeline for project tasks and their durations. Try the Gantt chart tool.
- Radar chart — plots several variables on spokes radiating from a centre, good for comparing profiles.
Work through the how to choose a chart decision guide, or settle a specific match-up in the chart comparisons — like pie vs bar or bar vs histogram.
Frequently asked questions
What are the main types of charts?
The everyday types of charts are bar (and column) charts for comparing categories, line charts for change over time, pie and donut charts for parts of a whole, area charts for cumulative totals over time, scatter plots for relationships between two numbers, and histograms for distributions. Stacked bar, grouped bar, radar and Gantt charts cover more specialised questions.
Is there a donut chart in the types of charts?
Yes. A donut chart (also spelled doughnut chart) is a standard chart type. It is a pie chart with the centre cut out, so it shows parts of a whole by slice angle and arc length, and the empty centre is often used to display the total or a headline figure.
Which chart is useful for comparing different categories side by side?
A bar chart. Each category gets its own bar on a shared axis that starts at zero, so readers compare lengths, which people judge accurately. If every category also has sub-groups, such as two years or two regions, use a grouped (clustered) bar chart so the sub-group bars sit side by side.
Which visualization shows the percentage of cancelled vs confirmed tickets: a pie chart, line chart, table or scatter plot?
A pie chart, because cancelled and confirmed are two parts of one whole that add up to 100 percent. A two-slice pie or donut makes the split obvious. A line chart only fits if you want to show how the cancellation rate changed over time, and a scatter plot is for relationships between two numeric variables.
How do I choose the right chart for my data?
Start from the question. Use a bar chart to compare categories, a line chart to show change over time, a pie or donut chart for a few parts of one whole, a scatter plot for the relationship between two numeric variables, and a histogram for how one variable is distributed. If readers need exact values, use a table.
Build your chart
Knowing the type is half the job — the other half is making one. Each maker is free, runs in your browser, and exports PNG or SVG with no signup or watermark.