A bar chart, also called a bar graph, compares numerical values across categories using rectangular bars. Each bar represents a category, and its length or height expresses the associated value on a shared scale. Readers use the bars to identify larger and smaller quantities, frequencies or other measures.
How a Bar Chart Works
On a vertical bar chart, categories occupy positions along the horizontal axis, while the vertical axis shows numerical values. This form is also commonly called a column chart. A horizontal layout reverses those roles: categories appear along the vertical axis, and values run along the horizontal axis. Horizontal bars can accommodate long category labels more comfortably.
Readers compare bar lengths from a common zero baseline. Positive values extend in one direction from zero, while negative values extend in the opposite direction. On a linear scale, a bar representing twice the value should extend twice as far from zero. Bar width normally does not encode the measure; bars generally have a consistent width so that length remains the meaningful visual cue.
Layouts With Multiple Series
A bar chart can also show more than one series for each category. Grouped bars place those series side by side within each category, allowing their values to be compared from the same baseline. The series need consistent identification, usually through colours and a legend or direct labels. Adding more series increases visual density rather than automatically solving overcrowding.
Stacked bars divide a category's total into component series, rather than treating every segment as a separate primary category. They show both totals and contributions to those totals. However, segments that do not share a baseline are harder to compare across bars. For stacks containing only positive values, each segment's length represents its component's value, and the full bar represents the combined total.
Looking for liquidity, exploring on-ramp/off-ramp services, or seeking expert guidance?
Get startedUses and Data Context
The main purpose of a bar chart is categorical comparison, such as comparing counts, amounts or percentages between named groups. A count shows how many observations fall into a category, while a percentage expresses a share of a stated total. Categories may follow a natural order or be sorted by value to make rankings easier to read. Sorting should not obscure an order that carries meaning.
Separate categories are different from the ordered numerical intervals used in a histogram. A histogram groups a numerical variable into intervals to show its distribution; it does not simply compare named groups. Similar-looking rectangles therefore do not necessarily represent the same kind of data.
Reading Values Accurately
A bar chart should use clear category labels, measurement units and a readable numerical scale. Identify the data source, reporting period and population or denominator so that readers know what the comparison covers. If error bars are included, check the stated measure of variability or uncertainty; error bars do not all represent the same quantity. Spreadsheet and visualization tools can create bar charts, but the author must still choose and label the data and scale appropriately. Check whether values are raw totals, rates or percentages, because each answers a different question. Where categories differ in size, raw counts alone may not support a like-for-like comparison.
Truncating the value axis away from zero can exaggerate differences by shortening the visible bars disproportionately. Data labels can give exact values, while gridlines can help readers estimate them; neither is essential if the scale is already clear. For grouped or stacked displays, confirm which series each colour represents before comparing values. The chart shows the reported differences, not why they occurred.