Line Chart Best Practices: 7 Rules That Fix Most Charts

The core line chart best practices are simple: use lines only for continuous data (usually time), plot four lines or fewer, pick a y-axis range that shows the trend honestly (a zero baseline is not required for line charts), and label lines directly instead of relying on a legend. Everything else is a refinement of those four rules.

When Should You Use a Line Chart?

Use a line chart when your x-axis is continuous and ordered, almost always time, and the point is how values change: stock prices, website traffic, temperature, monthly revenue. The connecting line implies that intermediate values exist between points, so if that implication is false, a line chart is the wrong choice.

• Good fit: continuous time series, trends, rate of change, comparing 2-4 related trends on the same axes
• Wrong choice: discrete categories (use a bar chart), parts of a whole (stacked area), distributions (histogram), correlation between two variables (scatter plot)

If you're unsure which chart fits your data, run through a chart selection framework first.

How Many Lines Should a Line Chart Have?

Three to four lines is the practical maximum; beyond five, the chart turns into spaghetti and readers stop tracing individual series. If you must show many series, highlight the one or two that matter in full color and render the rest in light gray, or split the chart entirely.

1-3 lines: ideal, easy to follow
4-5 lines: workable if colors are distinct
6+ lines:  spaghetti - restructure the chart

Fixes for many series:
- Small multiples (one mini chart per series)
- Highlight 1-2 lines, gray out the rest
- Interactive filtering or toggling

Whatever the count, make lines distinguishable: distinct colorblind-safe colors for 2-4 lines, varied dash patterns if the chart may be printed in black and white, and a thicker stroke for the primary series.

Should You Label Lines Directly or Use a Legend?

Label lines directly at their right endpoint whenever the layout allows it. A legend forces the reader's eye to bounce between the plot and a color key, and with similar hues that lookup fails entirely; a label sitting at the end of its line is read in a single glance. Reserve legends for cases where direct labels physically can't fit — many series ending at nearly the same value, or very narrow charts.

Legend (slower):            Direct labels (faster):

  ~~~~~~~~                     ~~~~~~~~ Product A
  --------                     -------- Product B
  ........                     ........ Product C
  ■ A  ■ B  ■ C

If two lines end too close together, nudge one label vertically or label at the line's peak instead. Matplotlib's ax.annotate, Plotly's mode="lines+text", and D3 all make endpoint labels a few lines of code.

Should You Ever Use a Dual-Axis Line Chart?

Almost never. Two independent y-axes let you make any two series appear correlated (or not) just by rescaling one axis, which is why dual-axis charts are the classic tool of misleading visualization. If the two series share units, plot them on one axis; if they don't, use two stacked panels with a shared x-axis, or index both series to 100 at the start.

The narrow legitimate case: two genuinely different units (revenue and temperature), both axes clearly labeled and color-matched to their lines, and an audience that expects the format. The full set of traps and the panel-based alternatives are in the dual-axis charts guide.

Does a Line Chart Need to Start at Zero?

No. The zero-baseline rule applies to bar charts, not line charts. Bars encode value by length, so truncating the axis literally lies about magnitude, but lines encode value by vertical position and trend by slope, neither of which requires zero to be honest. Forcing a zero baseline on data that lives far from zero (blood pressure in the 90-120 range, pH from 6.5 to 7.5, stock prices) flattens real variation into a useless straight line.

That said, axis range is still a rhetorical choice. Follow these rules:

• Include zero when absolute magnitude is the message (total revenue, units sold from launch).
• Crop the axis when small variations are the message, and pad the range so the data fills roughly two-thirds of the chart height.
• Never crop to exaggerate. A y-axis from $98M to $102M makes a 4% change look like a 10x swing. If you zoom, say so in a note or subtitle.
• Label the range clearly so readers can see the axis doesn't start at zero.

How Should You Handle Missing Data in a Line Chart?

Leave a visible gap in the line rather than silently connecting across the hole, because an unbroken line claims you know values you don't have. Most libraries do this by default when a value is null; interpolation should be an explicit, labeled choice, not an accident.

Missing data options, best to worst:
1. Break the line (gap) and note why data is missing
2. Dashed segment across the gap = "interpolated"
3. Impute values, clearly footnoted

Never: connect straight across a 3-month hole
       as if nothing happened

Related trap: uneven time intervals. Plotting Jan, Feb, Mar, Aug, Dec at equal spacing distorts every slope in the chart. Keep intervals consistent, and if you must skip periods, mark the break on the axis.

How Do You Annotate a Line Chart?

Annotate the two or three moments that explain the shape of the line: a product launch, a pricing change, a holiday spike. Raw trends rarely speak for themselves, and a short label at the right point saves a paragraph of caption text.

• Event markers with short labels ("Product launch", "Competitor entered")
• Shaded regions for periods ("30% growth phase", recession bands)
• Reference lines for targets, historical averages, or benchmarks (dashed)
• Confidence bands (shaded area) around forecasts to show uncertainty

Keep annotations subtle in weight so they support the data instead of competing with it. See chart annotation best practices for layout details.

What Are Small Multiples and When Should You Use Them?

Small multiples replace one crowded chart with a grid of mini line charts, one per series, all sharing the same axes and scale. They are the standard fix for the 6+ line problem: each trend stays readable, and shared scales keep the panels comparable at a glance.

Use them when series overlap heavily, when you have 5-20 series, or when the story is "which of these behave differently?" rather than "which is highest right now?" The one non-negotiable rule: identical y-axis ranges across panels, or comparisons become meaningless. Full walkthrough in the small multiples guide.

What Are the Most Common Line Chart Mistakes?

The most common mistakes are connecting discrete categories with a line, cramming in too many points or series, cropping the y-axis to exaggerate change, and mixing time granularities on one axis. Each one either lies about the data or buries the trend.

• Lines between categories: connecting Product A to Product B implies a continuity that doesn't exist. Use bars.
• Too much data: 10 years of daily points (3,650+) is noise. Aggregate to weekly/monthly or overlay a moving average.
• Misleading axis crop: zooming without saying so. Either widen the range or add a note.
• Mixed granularity: monthly and quarterly data on the same axis creates fake patterns. Convert to one granularity first.
• Marker overload: markers help below ~12 points (quarterly results), clutter above it (daily prices). Match markers to point density.
• Dual y-axes by default: two scales invite manipulation. Only use them for genuinely different units, clearly labeled.
• 3D effects: never. They add nothing and distort slopes.

How Do You Improve a Line Graph?

To improve a line graph, make the data lines the heaviest ink on the chart and lighten everything else: fade gridlines to pale gray, drop markers on dense series, replace the legend with direct labels, and rewrite the title so it states the takeaway. Then check the y-axis range and aspect ratio, because together they set how steep every slope looks.

When improving an existing chart, work through the edits in roughly this order; the early ones fix the most for the least effort:

1. Thicken the data, thin the scaffolding. Default styles often draw gridlines, axes, and borders at nearly the same weight as the data. Give the lines a heavier stroke and push everything else back. Horizontal gridlines are the reference lines that help readers match a point to its y-axis value, so keep them faint rather than deleting them when exact values matter; vertical gridlines can usually go.
2. Cut series to what the message needs. Keep color for the one or two lines the story is about and render context lines in light gray.
3. Swap the legend for direct labels at each line's right endpoint.
4. Rewrite the title as the insight ("Churn doubled after the price change") instead of a description ("Monthly churn rate"). See chart titles that explain.
5. Reset the y-axis range so the data fills most of the vertical space without exaggerating the change.
6. Fix the aspect ratio. A tall, narrow chart makes slopes look steeper than they are; a very flat one hides real movement. Wider than tall, around 2:1 or 3:2, suits most time series.
7. Annotate the turning points with one or two short event labels.

Steps 1-3 are the data-ink ratio principle applied to line charts: every pixel that isn't data should earn its place.

Quick Checklist Before Publishing

Run through this list before a line chart goes into a report or dashboard; each item takes seconds to verify and catches the mistakes readers notice most.

• ☐ Title states what's being shown
• ☐ Both axes labeled with units
• ☐ Y-axis range is honest (zero included or crop justified and labeled)
• ☐ Time intervals consistent; gaps shown, not interpolated silently
• ☐ No more than 4-5 lines (or small multiples used)
• ☐ Lines directly labeled with colorblind-safe colors
• ☐ Key events annotated
• ☐ Aspect ratio wider than tall (2:1 or 3:2)
• ☐ Source cited

← Back to Visualization Tips