How to Make Scatter Plot

Scatter plots show relationships between two numeric variables. They are ideal for correlation, calibration, dose-response, and measurement agreement figures.

Prepare your data

Use one column for the x value and one column for the y value. Add optional group, label, or replicate columns if you need color groups or annotated points.

Choose the visual encoding

Keep markers simple. Use color for meaningful groups, not decoration. If a trend line is included, label the model or keep the caption clear.

Label important points

Label only the points that matter: outliers, reference samples, top genes, or representative cases. Too many labels make the plot harder to read.

Common mistakes

Do not imply causation from correlation. Avoid heavy gridlines, oversized markers, and legends that repeat information already visible in labels.

Try it in SciFigure AI

Open the Scatter Plot Generator, upload paired data, and ask for a clean journal-style scatter plot.