How do I change an image from 72 DPI to 300 DPI?

In Photoshop: Image, Image Size, uncheck Resample, type 300 into Resolution. The pixel count stays identical, only the tag changes, and the announced print size is divided by about 4.17.

Does converting to 300 DPI improve quality?

No. The density tag holds no detail. Improving quality means adding pixels with plausible content, which only a neural upscale does properly.

My printer demands 300 DPI and my file says 72. Is that a problem?

Not if the file has the pixels. An image 7087 pixels wide labelled 72 DPI prints 24 inches at 300 DPI; the tag is corrected on export.

Guide · Resolution

Convert
72 DPI
to 300 DPI.

The manoeuvre takes ten seconds, it is written out below, and it will change nothing in your image. Here is why, and what to do instead when the file really is too small.

What you are actually changing

An image file holds a grid of pixels and, somewhere in the header, a line saying "when printing, pack them this many per inch". That line is the DPI. It occupies two bytes. It holds no detail, no colour, no visual information.

Going from 72 to 300 rewrites those two bytes. The pixel grid does not move. The direct consequence: the announced print size is divided by 4.17, since the same pixels are now packed four times tighter.

Worked example

	Before	After
Pixels	1920 × 1080	1920 × 1080
Tag	72 DPI	300 DPI
Announced size	26.7 × 15.0 in	6.4 × 3.6 in
Detail held	identical	identical

The image gained nothing and lost nothing. It simply started telling the truth about what it can print.

Step by step, per application

Photoshop

Image → Image Size. Uncheck "Resample", type 300 into Resolution, confirm. The Width and Height fields in pixels must stay frozen: if they move, the box is still ticked and Photoshop is manufacturing pixels by interpolation.

GIMP

Image → Print Size, then 300 in both X and Y resolution fields. Do not use "Scale Image", which resamples.

Preview, on macOS

Tools → Adjust Size. Uncheck "Resample image", type 300 into Resolution.

Without an editor

Drop the file into the checker: it reads the real tag and tells you immediately what size your pixels allow.

The real question: do you have the pixels?

A printer asking for "300 DPI" never means the tag. They mean enough pixels for the size ordered. The maths:

pixels = inches × 300

You want	You need	A 1920 px wide file gives
4 × 6 in	1200 px	enough
A4 / 8 × 10 in	2480 px	232 DPI, acceptable
A3 / 11 × 14 in	3508 px	164 DPI, poster
20 × 30 in	6000 px	96 DPI, no

If the row you care about is red, no tag manipulation will save you. You need pixels, and how you add them decides everything.

Adding pixels: two methods, two results

Interpolation (bicubic, Lanczos, Preserve Details) computes each new pixel from a weighted average of its neighbours. By construction it cannot invent anything: it stretches and softens. The result is larger and exactly as detailed as before, therefore blurrier at equal viewing distance.

A super-resolution network has seen millions of small-image / large-image pairs. It does not average, it proposes the most plausible content: a hard edge where interpolation would draw a gradient, a straw fibre where there was beige mush. The detail is not recovered, it no longer exists, it is rebuilt, and that is what shows on paper.

A caveat, measured rather than assumed: these networks smooth. On our reference case, the standard deviation of the Laplacian drops from 25.8 to 13.1 in raw network output. A sharpening pass at radius 1 pixel brings it back to 22.9. Without it, the print looks like wax, and a large-radius pass will not do: it only touches edges.

Check before you convert

Drop your file: real tag, real pixels, reachable size, format by format.

Check the file

Convert72 DPIto 300 DPI.