Passport photo guide

Rejected passport photo examples in 2026: the eight most common failure types, described precisely

Background is the top rejection reason — about 31% of cases. Head size errors follow at 24%. Shadows account for 18%. File format and size issues are about 14%. AI editing — now a fifth category — accounts for roughly 8% in 2026. Understanding what each rejection type actually looks like in a photo is the most reliable way to catch a problem before it becomes a hold notice three weeks after submission. This page describes each of the eight most common US passport photo rejection causes with enough precision that you can check your own photo against each description and identify whether any of these problems are present.

Download PassSnap See pricing

PassSnap guide

Capture · Verify · Download

Keywordpassport photo rejected examples

UpdatedSep 22, 2026

ReviewCrop, background, and AI verify

Example 1: shadow on the background (most common)

**What it looks like:** The area behind the head and shoulders in the photo is not a uniform white. Instead, there's a gradient — the background directly behind the head is white or close to white, but the background behind the shoulders, neck, and outer edges of the head transitions to a noticeably darker grey or off-white. The shadow is typically shaped like a halo or shadow portrait — it follows the silhouette of the head and upper body. **Why it fails:** The State Department's background uniformity check compares the brightness values of pixels in the background area. A uniform white background has consistent pixel values. A shadow gradient has pixel values that decrease progressively from the lighter centre to the darker edges. The automated check identifies this gradient as a non-uniform background. **What to change:** Move at least one metre from the background wall before taking the photo. At close range (less than half a metre), the head casts a visible shadow directly on the wall behind it. At one metre or more, the shadow falls below and behind the camera's field of view. Also turn off overhead ceiling lights and use front-facing natural light — ceiling lights create downward shadows that fall on the background wall behind the shoulders.

Example 2: shadow on the face

**What it looks like:** The face has uneven brightness distribution. The most common form is a shadow in the eye sockets — the area under the brow ridge and above the cheekbones is significantly darker than the rest of the face. The nose often casts a visible shadow downward onto the upper lip area. The chin area may fade into shadow below the jawline. In more pronounced cases, one side of the face is noticeably brighter than the other. **Why it fails:** Biometric facial recognition systems use the brightness distribution across the face to estimate the positions of facial landmarks — the distance between the eyes, the width of the nose, the position of the mouth corners. Shadows in the eye socket area register as the eyes being more deeply recessed than they actually are. The resulting biometric template doesn't match the applicant's actual face geometry. **What to change:** Turn off overhead ceiling lights entirely and use front-facing natural daylight from a window as the primary light source. A window directly in front of the subject at or slightly above face level eliminates the downward shadow that ceiling lights create. If natural light isn't available, use a ring light at face height — not positioned above the head.

Example 3: head too small (most common compositional error)

**What it looks like:** In the 2×2 inch crop, there is a significant amount of white space above the head and below the chin. The face occupies less than half the vertical height of the frame. The top of the head may be near the middle of the frame rather than near the top. The shoulders are prominent in the lower portion of the frame. **Why it fails:** The head must measure between 1 and 1⅜ inches from chin to crown — 50 to 69 percent of the 2-inch frame height. When the face occupies less than 50 percent of the frame height, the photo fails the automated head size measurement check. **What to change:** Move the camera closer to the subject. At 2 to 2.5 metres, the face is typically too small in the 2×2 crop. At 1.5 metres, the face is typically within the correct range for most adults. Take a test shot, crop it to the 2×2 proportions in the camera roll, and check whether the face reaches at least 50 percent of the frame height before committing to the full session.

Example 4: head too large or too tightly cropped

**What it looks like:** The face fills more than 70 percent of the 2×2 frame height. The crown of the head is at or very close to the top edge of the frame, leaving almost no space above the head. Alternatively, the ears and sides of the head are cut off at the frame edges. **Why it fails:** The automated head size check has an upper boundary as well as a lower one. Above 69 percent of frame height, the photo fails the measurement check. Additionally, cutting off the sides of the head, the top of the head, or the ears fails the "full face visible" requirement. **What to change:** Step further back from the camera, or switch from the telephoto lens to the primary lens (if you were using a zoom setting). Leave visible space above the crown of the head — approximately the width of the forehead — and confirm both ear edges are within the frame.

Example 5: glasses visible

**What it looks like:** The subject is wearing glasses in the photo. This includes prescription glasses, reading glasses, sunglasses, or any other eyewear. The glasses are fully or partially visible in the photo. **Why it fails:** Glasses have been prohibited in US passport photos since November 2016. The prohibition covers all glasses regardless of prescription strength or frequency of use. Glasses can obscure the eyes, create glare on lenses, and alter the apparent shape of the eye area — all of which affect biometric accuracy. **What to change:** Remove glasses before the session. If you cannot remove glasses for a documented medical reason, the only exception is a signed statement from a medical professional confirming the medical necessity, submitted with the application.

Example 6: AI-processed background (2026-specific)

**What it looks like:** The background appears uniform white, but the edges of the subject — particularly around the hair, ears, and shoulder contour — show slight irregularities. The transition from the subject to the background may look too sharp and clean, lacking the natural softness that exists where a person stands against a background in real photography. Around loose hair strands, the area that should be a natural mix of hair and background often shows a clear, sharp cutoff. Occasionally, a slight colour halo or fringe appears at the subject boundary. **Why it fails:** AI background replacement tools use semantic segmentation — an algorithm that classifies each pixel as "person" or "background" and then replaces the "background" pixels with white. The boundary between classified regions carries statistical artifacts that differ from natural photographic subject-to-background transitions. The State Department's automated review system identifies these artifacts as evidence of AI image alteration. **What to change:** Use a real white background in the photo session rather than replacing a non-white background digitally. A white sheet, white wall, or white foam board photographed under front-facing neutral light produces a background that doesn't carry AI processing artifacts. The background must be correct in the original photo — there's no compliant post-capture fix.

Example 7: AI skin smoothing or Portrait mode (2026-specific)

**What it looks like:** The face appears unusually smooth, with reduced visible texture in skin areas compared to the hair and clothing in the same photo. The boundary between the face and background is either very sharp (AI-generated depth segmentation from Portrait mode) or the background appears progressively blurred in a way that follows a depth map rather than the natural optical blurring that occurs at camera distance. **Why it fails:** AI skin smoothing produces a statistical discontinuity — skin areas have lower noise levels than non-skin areas in the same photo, which is a signature of AI-based alteration. Portrait mode background blur is generated by a depth-map algorithm rather than by natural optical defocus, and the depth-map pattern is statistically distinguishable from optical blur. **What to change:** Disable Photographic Styles on iPhone (set to Standard), disable Samsung Beauty mode and Scene Optimizer, disable Google Pixel's Face Unblur and Top Shot. Use standard Photo mode, not Portrait mode, on any phone. Do not apply any skin smoothing or beauty filter after capture.

Example 8: wrong expression or head tilt

**What it looks like:** The mouth is open with teeth visible; or the expression is wide enough that the cheeks are raised significantly from their neutral position; or the eyes are narrowed from squinting or laughing; or the head is tilted to one side; or the chin is raised or lowered so that the face is no longer parallel to the camera. Hats, headbands, and non-exempt head coverings are also in this category. **Why it fails:** Open-mouth expressions and wide smiles change the apparent positions of biometric landmarks. Raised cheeks from a wide smile narrow the visible eye area and change the cheek width. Head tilt, raised chin, or lowered chin change the face geometry in ways that affect matching accuracy. **What to change:** Aim for a relaxed, neutral expression — jaw slightly relaxed, mouth closed, eyes open naturally (not widened), face level with the camera. The exhale technique helps: breathe in, exhale slowly, and open the shutter at the end of the exhale, when the face naturally settles to its resting position.

How PassSnap fits

PassSnap's optional AI verify step checks the exported photo for the main verifiable rejection categories: background uniformity (catches example 1, and flags non-uniform edges), expression and eye openness (example 8), and glasses detection (example 5). The guided capture addresses example 2 (face shadow — through real-time framing feedback), example 3 and 4 (head size — through real-time proportion feedback), and example 6 and 7 (AI processing — no AI enhancement is applied to the export). For the complete pre-submission check, reviewing the exported crop at full resolution against the pre-submission checklist is the final step.

The pre-submission checklist

Background is uniformly white with no grey gradient behind the shoulders or head.

Face occupies 50 to 69 percent of the frame height — not less, not more.

Eyes are fully open, both visible, and equally lit.

No shadow in the eye sockets, under the nose, or under the chin.

No glasses (prescription, reading, sunglasses, or tinted lenses).

Mouth is closed in a neutral, relaxed resting expression.

Head is level — no tilt, no raised or lowered chin.

Background was photographed in the session, not digitally replaced with AI.

FAQ

What does a rejected passport photo look like?

The most common types are: background with a shadow gradient behind the head and shoulders (too close to the wall); face occupying less than 50 percent of the frame height (camera too far away); uneven face lighting with shadows in the eye sockets (ceiling lights as primary source); AI-replaced background with unusual edge characteristics at the subject boundary (background replacement tool used); and Portrait mode or AI skin smoothing applied before submission. Each of these produces a photo that looks correct to the eye but fails the automated review system.

What's the most common reason a passport photo is rejected?

Background errors are the most common rejection cause — about 31% of rejections. Specifically, shadows on the background caused by standing too close to the wall, and backgrounds that aren't sufficiently white due to warm lighting. Head size errors (24%) and shadows on the face (18%) are the second and third most common, followed by file format and size issues (14%) and AI processing artifacts (8%, a new category in 2026).

Can I fix a rejected passport photo?

File format and size errors can be fixed by re-exporting the original photo in the correct format. Background, lighting, AI processing, expression, and head size errors require retaking the photo. You cannot edit out a shadow, replace a non-white background, or de-process an AI-altered photo and have the result be compliant. The fix for visual problems is always a new photo session with the problem corrected at the source.