How to Make a Passport Photo the Right Size
Short answer
Get the exact spec from your own issuing authority's published page — dimensions and head size vary by country and change. Convert it to pixels: centimetres ÷ 2.54 × the DPI. Then crop to the required ratio first, resize to the exact pixel numbers, and compress to any stated KB range last.
On this page
Passport and visa photo requirements get treated as a resizing problem, and the resize is the easy part. The specification is where people come unstuck, because it varies by country, varies by document type within a country, and changes.
So the first instruction is unglamorous: get the spec from the issuing authority’s own page. Not from a photo booth’s website, not from a forum post, not from this guide. Your passport office, embassy or visa service publishes the exact dimensions, head size and background requirements, and that published page is the only version that counts. Everything below is the method for meeting whatever it says.
What the specification will tell you
Expect four or five separate requirements, each checked independently.
| Requirement | Typically expressed as | Why it fails |
|---|---|---|
| Physical dimensions | Centimetres or inches | Wrong aspect ratio after cropping |
| Resolution | DPI, where printed photos are involved | Confused with pixel count |
| Head size | A measurement, or a proportion of the height | Cropped by eye |
| File size | A KB range for digital uploads | Compressed too hard or too little |
| Background and lighting | Described in words | Shadows, wrong shade, patterned wall |
The head size requirement is the one most often missed, and automated checkers measure it precisely. A photo at exactly the right pixel dimensions with the head too small in the frame is rejected, and the error message rarely says so.
Converting the spec into pixels
If the requirement is in centimetres with a DPI figure, it is describing a pixel count in a roundabout way. Two steps:
centimetres ÷ 2.54 = inches
inches × DPI = pixels
Worked through for a common pattern:
3.5 cm ÷ 2.54 = 1.378 inches
1.378 × 300 DPI = 413 pixels
If the measurement is already in inches, skip the first step. A 2-inch square at 300 DPI is 600 × 600 pixels.
Do the same arithmetic for the head size requirement. If the spec says the head must be a given height, convert that to pixels too, and you then have a concrete target to crop against rather than a visual judgement.
The full version of this conversion, including why the DPI tag by itself changes nothing, is in pixels, cm and DPI explained and what 300 DPI actually means.
Taking the photograph
If you are shooting it yourself, the capture decides whether the rest is possible.
Background. A plain, light, unpatterned wall. Stand about a metre in front of it rather than against it, which prevents a shadow appearing behind your head. Shadows behind the subject are one of the most common automated rejections.
Lighting. Even, indirect, from the front. Daylight through a window on an overcast day is close to ideal. Avoid direct sunlight, a single lamp to one side, and overhead lights that cast shadows into the eye sockets.
Camera position. At eye level, roughly an arm’s length and a half away. A phone held close distorts facial proportions noticeably, and a photo taken from below or above fails the framing rules.
The subject. Face square to the camera, eyes open and visible, neutral expression with the mouth closed unless the spec says otherwise, nothing covering the face. Glasses are increasingly disallowed outright by many authorities; where permitted, there must be no glare and the eyes must be fully visible. Check what yours says rather than assuming.
Resolution. Shoot at full quality. You will crop heavily, and cropping discards pixels. Starting from a 12-megapixel photo leaves plenty; starting from a screenshot of a video call does not.
Have someone else take it. A selfie at arm’s length is too close and the proportions show it.
Crop to the ratio, then resize
This order matters and getting it backwards is the most common technical mistake.
Your photo is 4:3 or 3:2. The requirement is something else — a 3.5 × 4.5 cm spec is 7:9, which is considerably taller relative to its width. Resizing straight to the target dimensions squashes the image, and a stretched face on an ID photo is immediately obvious to a reviewer.
- Crop to the ratio. Most crop tools accept a custom ratio, or you can crop to the exact target pixel dimensions directly if your tool allows it. Position the crop so the head size and position match the spec.
- Resize to the exact pixel numbers from your conversion above.
- Compress if a KB range is specified.
Platform notes:
- Windows Paint: crop first, then Resize → Pixels, and untick Maintain aspect ratio only after the crop already matches the target ratio.
- macOS Preview: crop with the selection tool, then Tools → Adjust Size with units set to pixels.
- Phone: the Photos app crop tool on both platforms supports custom ratios. Crop there, then resize on a computer or with a browser-based tool.
Our image compressor handles the resize and the file size in one step and runs in your browser, which matters for an identity document — a passport photo has no business being uploaded to an unknown server for processing.
Hitting a file size range
Digital applications frequently specify both a minimum and a maximum in kilobytes. Both directions can fail.
Too large is the familiar problem; compressing a photo to an exact size in KB covers hitting a target without blurring the face.
Too small catches people out after an aggressive compression. The honest fix is to re-export at higher quality rather than padding the file, and increasing an image file size to meet a minimum covers it.
Save as JPG unless the spec says otherwise. It is the format every validator accepts. If your source is an iPhone HEIC, convert it with our HEIC to JPG converter before anything else — a format rejection often surfaces as a vague dimension error, which sends people resizing a file that was never the problem.
Checking before you submit
Go through the authority’s own list, item by item, with the finished file open at 100% zoom:
- Pixel dimensions match the converted numbers.
- Head height matches the stated measurement.
- Background is plain, the right shade, and free of shadows.
- Eyes visible, expression neutral, nothing obscuring the face.
- File size inside the stated range.
- Format matches what is asked for.
Many application portals include a photo checker before final submission. Use it. It is the same validator that will reject the application later, and it costs nothing to fail it now.
What to expect
With the specification in front of you, the technical work takes about ten minutes: shoot, crop to ratio, resize, compress, check.
The honest limit is that the photograph itself is harder than the file. Framing, lighting and background account for most rejections, and no amount of resizing fixes a shadow behind the head or a head cropped too small. If a home attempt fails twice, a booth or a photographer who does this routinely is usually the cheaper option in time if not in money. And if the rejection message is vague, work down the authority’s checklist rather than assuming it was the dimensions — why photo uploads get rejected covers separating the four possible failures.