Western blot densitometry, step by step

About 8 min read · Updated 2026-09-30

Densitometry turns “this band looks darker” into a number. It comes down to four steps: measure band intensity → subtract background → divide by the loading control → set the reference lane to 1.

Try it with the tool

Upload a blot image — lanes and bands are found and quantified automatically, with replicate statistics and a publication figure and slides.

Open the western blot tool →

1. Prepare the image

Use the original TIFF saved by the imager; JPEGs and screenshots change values through compression and adjustment.
Never quantify an image whose brightness/contrast was adjusted — apply cosmetic adjustments only to a copy for the figure.
Choose an exposure in which bands are not saturated (common quantification mistakes).

2. Band intensity: integrated density

Integrated density is the sum of pixel signal inside the box (ROI) around a band. For dark bands on a white background, convert each pixel to signal = 255 − brightness (8-bit) before summing.

Integrated density = Σ (pixel signal − background)

Measure all lanes of one protein with boxes of the same size for a fair comparison.

Plot intensity along the lane and each band becomes a peak. The shaded area — the area under the peak (integrated density) minus the background below the dashed line — is the true band signal.

3. Background subtraction

Method	How	Pros / cons
Box border (local)	Median of the pixels on the band box edge	Adapts to lane-to-lane background; recommended in most cases
Empty region	Mean of a separate box drawn where there are no bands	Simple if the background is even; inaccurate with smears
Rolling ball	Estimates and removes smooth background across the image (ImageJ)	Handles diffuse background; values depend on the radius

State the method in your Methods so others can reproduce it.

4. Normalization and relative values: example

A textbook example: p-ERK after EGF stimulation, normalized to β-actin (background-subtracted integrated density).

Lane	p-ERK	β-actin	Ratio	Relative
Control	12,000	40,000	0.300	1.00
EGF 5 min	30,000	38,000	0.789	2.63
EGF 15 min	52,000	41,000	1.268	4.23

Ratio = target ÷ loading control · Relative = ratio ÷ ratio of the reference lane

Phosphoproteins are often normalized to the total form of the same protein (e.g. p-ERK / total ERK) instead of a loading control.

The relative values from the table above. With the control set to 1, EGF gives 2.63× at 5 min and 4.23× at 15 min.

5. Replicate statistics

Exposure and sensitivity differ between blots, so raw values cannot be compared across blots. Pool the relative values (reference lane = 1 on each blot) and report mean ± SD; n is the number of independent blots.

Because the reference is always 1, test “is the treatment different from 1?” with a one-sample t-test, or use ANOVA for several groups. Log-transforming the ratios before testing is also common.