How to Read a Histogram and Understand Its Shape

A histogram is a chart for understanding the distribution of numerical data, such as test scores, travel times, or package weights. Its bars show how frequently values occur within consecutive ranges. Reading one helps you see where observations cluster, how widely they vary, whether the data are lopsided, and whether any areas seem unusual. Before you draw conclusions, check what was measured, the units, and how the intervals were chosen.

What You’ll Need

A histogram to inspect, either printed or on a screen.
A basic understanding of the variable being measured and its units, such as minutes, inches, or points.
If you are making a histogram from raw data: the numerical observations and a spreadsheet, calculator, or graphing tool. These tools are optional when you are only reading an existing chart.

Step-by-Step

Identify the variable. Read the chart title, labels, and any caption. Determine what each observation represents and whether the data describe a sample, a group, or a period of time. A chart of people’s ages means something different from one of daily temperatures, even if the shapes look alike.
Read the horizontal axis. This axis lists numerical intervals, also called bins or classes. Check their boundaries and units. For example, a range from 10 to 20 might mean measurements at least 10 but less than 20; the chart or its notes should clarify boundary conventions when they matter. Notice whether bin widths are equal. If they are not, bar area—not height alone—may represent frequency.
Read the vertical axis. Find out whether bar height represents a count, a percentage, a proportion, or a density. A count tells how many observations fall in a bin; a percentage gives the share of the total. For unequal-width bins, a density scale may be used so that bar area represents frequency. Do not assume the axis shows counts just because the bars look familiar.
Choose a bar and interpret it. Match the bar’s horizontal range to its vertical height. If the vertical axis is labeled “number of observations,” a bar reaching 8 means eight observations are in that interval. If the axis shows percentages, interpret the height using that scale instead. Each bar covers a range, so the chart usually cannot tell you the exact value of an individual observation.
Find where values concentrate. Look for the tallest bar or group of tall neighboring bars. This shows the interval or intervals containing the most observations, often called the peak or mode region. A histogram can have one main peak, several peaks, or no clear peak. A peak identifies a busy range, not necessarily a single most common exact value.
Describe the overall shape. Check whether the distribution is roughly symmetric, skewed, or multi-peaked. In a roughly symmetric histogram, the sides have broadly similar shape around the center. In a right-skewed histogram, a thinner tail extends toward larger values; in a left-skewed one, it extends toward smaller values. Two or more separated peaks may suggest distinct subgroups, but the chart alone does not prove why they differ.
Assess the spread and center. Note the range covered by the bars and where the bulk of observations lies. A wider span indicates more variation in the displayed values, while a narrow cluster indicates less. You can estimate a typical region from the main cluster, but a histogram does not give an exact mean or median unless those are separately marked or calculated from the data.
Check for gaps and unusual bars. Empty intervals, isolated bars, or a long tail can signal unusual values or separate groups. First confirm that the gap is real and not caused by the chosen bin boundaries, missing observations, or a very small sample. An unusual bar is a reason to inspect the source data, not proof of an error.
Summarize in context. State what the histogram shows using the variable and units: where values cluster, how the distribution is shaped, and any notable spread or gaps. Keep claims proportional to the display. A histogram describes the data shown; by itself, it does not establish a cause, explain how data were collected, or guarantee what will happen next.

Common Mistakes

Treating bars like categories: Histogram bars represent ranges of numerical values. Their order matters, and adjacent bars usually touch because the intervals are continuous. A bar chart, by contrast, compares separate categories and usually has gaps.
Ignoring the vertical scale: A tall bar may represent a count, percentage, or density. Read the label and units before comparing heights, especially when bins have different widths.
Assuming each bar is one value: The observations are grouped. You cannot generally recover the exact data points, or an exact mean, from the visible bars alone.
Calling every uneven distribution an error: Skew, multiple peaks, or a long tail can be genuine features. Check the data and context before deciding something is wrong.
Comparing histograms with different bins or scales: Different interval widths, axis limits, or vertical scales can make the same data look different. Check that the displays are comparable before drawing conclusions.

Tips

Read the title and axis labels before interpreting any bar; units can change the meaning completely.
Use approximate language such as “most values fall between” or “the distribution has a longer upper tail.” Avoid claiming exact values that the grouped chart cannot show.
When comparing groups, look for differences in center, spread, shape, and sample size. If counts are shown and groups contain different numbers of observations, percentages or another normalized scale may give a fairer comparison.
If the bin size is adjustable, try another reasonable bin width. Very wide bins can hide structure; very narrow bins can make random variation look like a pattern. The underlying data have not changed when the display changes.
For a fuller summary, consult the raw data or accompanying statistics. A histogram is useful for seeing distribution shape, but it is not a substitute for every numerical measure.

Troubleshooting

The bars look backward or confusing: Check whether the horizontal axis increases from left to right and whether the chart has been rotated or labeled unusually. Follow the printed scale rather than guessing from appearance.

The vertical axis has unfamiliar values: Look for words such as “percent,” “relative frequency,” or “density,” and check whether the axis begins at zero. If it shows density, compare bar areas when bin widths differ.

You cannot tell what the bars mean: Look for a caption, legend, notes, or data source explaining the variable, units, bin boundaries, and treatment of missing values. If these details are absent, describe only what is visible and treat more specific conclusions as uncertain.

The histogram seems to contradict a reported average: The average can be pulled toward a long tail, while the tallest bars show where observations are most concentrated. These describe different aspects of the distribution and can both be correct.

A gap or isolated bar seems suspicious: Check the original observations, how measurements were recorded, and whether the chart uses narrow bins. Do not remove or recategorize values without a documented reason.

Frequently Asked Questions

What does each bar in a histogram represent?

Each bar represents observations within a numerical interval. Its height or area indicates frequency, depending on the vertical scale and bin widths.

Why do histogram bars usually touch?

The bars represent neighboring intervals on a continuous numerical scale, so there are no category gaps between them. Gaps can appear where an interval contains no observations.

What does a right-skewed histogram mean?

Most values are toward the lower end, with a tail extending toward larger values. Skew describes the shape; it does not by itself explain the reason for that shape.

Can I find the mean from a histogram?

Not exactly in most cases, because values have been grouped into intervals. You can estimate a center, but calculating the exact mean requires the original observations or additional information.

Does a histogram show outliers?

It can reveal isolated bars or a long tail that may contain unusual values. The bins may hide how far apart individual observations are, so use the raw data to check possible outliers.

How is a histogram different from a bar chart?

A histogram displays numerical data grouped into ordered intervals, usually with touching bars. A bar chart compares distinct categories, whose bars are generally separated.

Further reading: NIST/SEMATECH e-Handbook of Statistical Methods. Details can change, so check the official page for anything that matters to you. Spotted a mistake? Tell us. How we write an answer.