PSA Grading Standards Explained: The Real Tolerances at Every Grade
PSA publishes a grading scale, but the published descriptions are vague enough that collectors regularly misread their own cards. This guide is the actual working version: the tolerances PSA graders apply, the cutoff between each grade, and the specific flaws that drop a card from a 10 to a 9 to an 8.
Everything below is what an AI grading app like ZeroPop measures against, and what PSA's human graders look for under their loupes.
The PSA 1–10 Scale, Decoded
PSA 10 (GEM-MT). Effectively flawless. Corners must be sharp under loupe inspection. Edges with no visible whitening. Surface with no scratches, scuffs, or print defects. Centering must be no worse than 55/45 left/right and 60/40 top/bottom on the front (60/40 and 75/25 respectively on the back). PSA 10s are roughly 1 in 5 to 1 in 20 of submissions for modern cards depending on the product.
PSA 9 (MINT). One minor flaw. PSA 9s typically have one of: very slight corner wear under magnification, micro-whitening on one or two edges, a single tiny print dot, or centering between 55/45 and 60/40 front. Multiple minor flaws drop the card to PSA 8. Most "perfect-looking" modern cards fall into PSA 9 territory.
PSA 8 (NM-MT). A handful of minor flaws or one moderate one. Light corner wear visible to the naked eye. Edge whitening on two or more edges. A faint scratch on the surface. Centering up to 65/35 front and 80/20 back. Most well-kept cards from packs fall into PSA 8 if the centering is off.
PSA 7 (NM). Clear, but one or two notable flaws. Visible corner wear. Edge wear on multiple edges. Surface with a small scratch or scuff that catches light. Centering up to 70/30 front. Cards in PSA 7 territory are usually not worth the submission cost unless the card itself is very valuable.
PSA 6 (EX-MT). Played condition with multiple visible flaws. Soft corners, edge wear, surface marks visible at arm's length. Used to be common for vintage; rare for modern submissions because the math doesn't work below PSA 7 for most cards.
PSA 5 (EX). Played card with rounded corners, surface scratches, and possibly minor creasing. Vintage-only territory in practice.
PSA 4 (VG-EX) through PSA 1 (POOR). Severe wear, soft corners, visible creases, surface damage. PSA 4 and below is almost exclusively for vintage cards with collector value despite condition.
The Four Sub-Grade Categories
PSA grades on four categories internally, even when the slab only shows one number. (BGS publishes all four; PSA does not, but uses them.) Knowing what each one measures lets you predict your grade accurately.
Corners. PSA inspects corners under loupes for sharpness, whitening, and dings. A corner is considered "sharp" if it forms a clean point with no rounding. White on a dark-bordered card is the most common giveaway. Even microscopic whitening drops the corner score.
Edges. The four edges of the card. Whitening, chipping, and rough cuts move the score down. Modern cards with print-line whitening at the cut line are common. This is a manufacturing flaw but PSA grades the card as it arrives.
Surface. The face of the card. Scratches, scuffs, print defects (color shift, banding, dimples), and gloss issues all count. Holographic surfaces are inspected at multiple angles because some flaws only catch light at 45°.
Centering. The single most cited reason for PSA 9s instead of PSA 10s. Measured as left/right and top/bottom border ratios. Even visually perfect cards can fail centering and drop a grade.
PSA Centering Tolerances (The Actual Numbers)
This is where most submissions fail. Here are the actual PSA centering tolerances by grade:
PSA 10: 55/45 or better front; 75/25 or better back.
PSA 9: Up to 60/40 front; up to 90/10 back.
PSA 8: Up to 65/35 front; up to 90/10 back.
PSA 7: Up to 70/30 front; up to 90/10 back.
The back tolerance is more forgiving because back centering is less visible, but extreme back miscentering still drops a grade. Note: BGS centering tolerances are tighter than PSA. A card that gets PSA 10 centering might only get BGS 9.5 centering. We cover this in BGS sub-grades explained.
The Math of "Lowest Sub-Grade Wins"
PSA does not publish their formula, but the working model that matches outcomes is: the lowest sub-grade anchors the overall grade, with a small bump if the gap to the second-lowest is large enough.
If your sub-grades are corners 9, edges 10, surface 10, centering 10, you're getting a PSA 9. Not a 10. The 9 in corners drags everything down.
If your sub-grades are 9.5, 10, 10, 9.5, the floor is 9.5 and you'll get a PSA 10 (PSA does not have half-grades on the slab; 9.5 rounds to 10 in their system).
If your sub-grades are 9, 9, 9, 9, you're getting a PSA 9. Never a PSA 10, because no sub-grade reaches 10.
This is exactly the formula ZeroPop's grading engine uses to convert sub-grades into a final overall prediction. It's also why centering matters so much: it's the most common sub-grade that drops below 10.
What Gets a Card to PSA 10
For a card to clear PSA 10, all four sub-grades need to be at 10. Practically, that means:
- Corners: sharp under loupe inspection, no whitening, no rounding.
- Edges: clean cuts, no whitening on any edge.
- Surface: no scratches, scuffs, print dots, or holo damage.
- Centering: within 55/45 front and 75/25 back.
Cards that meet all four are rare. For most modern Pokémon and sports products, PSA 10 rates fall between 5% and 25% of submissions. Pre-screening is the difference between submitting blind and submitting profitably.
Common Reasons Cards Drop From 10 to 9
In order of frequency:
- Centering off by a few percentage points. The card visually looks centered but measures 56/44, just over the PSA 10 threshold. By far the most common cause.
- Single corner with micro-whitening. Often invisible to the naked eye but flagged under loupes.
- Print line on the surface. Manufacturing flaw, not your fault, but counts.
- Edge whitening on the cut line. Cards from sheets where the cut introduced micro-whitening fail PSA 10 even at the bottom edge.
- Holo surface scratch from sleeve insertion. Common on freshly pulled holos that were sleeved incorrectly.
How to Pre-Screen Like a PSA Grader
Before submitting any card, do this checklist:
- Loupe the corners under bright light. White spots = not a 10.
- Tilt the card at 45° to the light. Look for surface scratches and dimples.
- Measure centering numerically. Eyeballing it lies. An AI grading app gives you the L/R and T/B ratios.
- Inspect all four edges for whitening and chipping.
- Check the back for centering, surface, and edge issues. The back can drop a grade.
If anything fails this checklist, the card is likely a PSA 9 or below. Use the submission ROI math to decide whether the submission is profitable at PSA 9.
How AI Card Grading Apps Apply These Standards
A modern AI card grading app applies the same rubric PSA uses, but it does it from a phone photo. The four sub-grades are scored against the same tolerances. The centering ratios are measured to two decimal places. More precise than the human eye.
The win is consistency: the same card scanned twice returns the same grade. That's something even PSA cannot guarantee on borderline cards. AI is not better than PSA at the platonic ideal of "what is this card's grade." AI is better at telling you whether your specific card clears the PSA 10 threshold reliably enough to be worth the submission fee.
Frequently Asked Questions
What is the PSA 10 centering tolerance?
55/45 or better on the front (left/right and top/bottom) and 75/25 or better on the back. Anything outside these bounds drops the card to PSA 9 even if every other sub-grade is perfect.
How does PSA decide between a 9 and a 10?
PSA grades each card on four sub-categories. Corners, edges, surface, and centering. The lowest sub-grade anchors the overall. If any sub-grade is below 10, the card cannot be a 10. Centering is the most common reason a visually perfect card scores 9.
Are PSA grading standards the same as BGS or CGC?
No. The categories are similar but the tolerances differ. BGS centering is tighter at the gem level. CGC is more forgiving on micro-whitening. We cover the differences in PSA vs BGS vs CGC.
Can a card get PSA 10 with one corner slightly soft?
No. PSA 10 requires sharp corners under loupe inspection. Even a single soft corner drops the grade to PSA 9. The lowest sub-grade always wins.
What's the easiest way to predict my PSA grade before submitting?
Use an AI card grading app to measure centering ratios numerically and check sub-grades for corners, edges, surface, and centering. The best card grading apps report these in the same format PSA uses internally, so the prediction maps directly to the human grade.
Why does PSA give different grades for the same card on resubmission?
PSA grading involves human judgment on borderline cases. A card right on the 9/10 line may grade differently across submissions. This is why pre-screening matters. Submit only cards that are clearly above the PSA 10 threshold on every sub-grade, not borderline candidates.
Written by
Marcus ReevesLead Grading Editor, ZeroPop
Marcus has been collecting and grading trading cards since the late 1990s, with a focus on Pokemon, vintage baseball, and modern basketball. He leads ZeroPop's grading research, runs the editorial team's PSA, BGS, and CGC submission tests, and writes the cost and turnaround tracking that powers the app's ROI calculator.
Know your grade before you submit.
ZeroPop scans your cards and gives instant sub-grades for corners, edges, surface, and centering. PSA, BGS, and CGC estimates included. Free to start.
Get Free on iOSKeep reading
BGS Sub-Grades Explained: How Beckett Calculates the Final Grade
How BGS sub-grades for centering, corners, edges, and surface combine into a final grade. Including the BGS 9.5 Gem Mint and BGS 10 Pristine thresholds.
Grading Company ComparisonsPSA vs BGS vs CGC vs SGC: Complete Grading Company Comparison 2026
An honest, detailed comparison of every major card grading company (PSA, BGS, CGC, and SGC) covering cost, speed, and market value.
Grading GuidesIs My Card Worth Grading? How to Decide Before Spending Money
A practical framework for deciding whether to grade your cards. Covers the break-even math, common traps, and the AI pre-grading approach.