DAT PAT Score Percentiles: 20 vs 22 Explained

A raw PAT score means almost nothing by itself: what actually matters is the percentile behind it. On the old 1–30 scale, a 20 typically sits around the mid-70s percentile, while a 22 sits closer to the mid-80s, which is a much bigger jump than "two more points" sounds like. Once you see the PAT score percentile chart laid out, you stop chasing an arbitrary number and start targeting the percentile your actual target schools expect. But hitting that percentile isn't a knowledge problem: the PAT is a trainable visual skill, and the thing that actually moves your percentile is the number of genuinely new questions you rep against, not how many times you've memorized the same few hundred.

We scored 97th-plus percentile on the DAT ourselves (legacy 25 AA with a 30 in organic chemistry, and 27 AA with a 29 TS) and now attend the #1 dental school in the world. We got there by knowing exactly what percentile we needed and building toward it: that's the same shift we want for you here.

What the PAT Score Percentile Chart Actually Shows

The PAT is scored and reported completely separately from your Academic Average. It never touches your Bio, GC, OC, RC, or QR scores. It's its own number, judging spatial reasoning through six subsections: keyholes, top-front-end, angle ranking, hole punching, cube counting, and pattern folding.

A percentile tells you the percentage of test takers you scored above. A PAT percentile of 80 means you outscored roughly 80 out of 100 people who sat for the exam: a completely different piece of information than the raw scaled score, and the number admissions committees actually care about.

Because DAT scores follow something close to a bell curve, most test takers bunch up near the middle. Small raw-score differences near the average swing your percentile a lot; the same size difference out on the tail barely moves it.

PAT Score Percentile Chart: Legacy 1–30 Scale

Here's the general shape of the PAT score percentile chart on the legacy 1–30 scale, which most forum threads and older guides still reference. Treat these as close approximations, not exact figures: percentile ranks shift slightly year to year, so check the ADA's own current table for the precise number tied to your test date.

Since March 2025, the DAT reports PAT on a new 200–600 scale in 10-point increments, with roughly 400 sitting near the middle of the pack (47th–54th percentile per the ADA’s official table), about the 50th percentile, the same anchor point as the old 17. Every 10-point step above 400 moves your percentile in the same general direction as the legacy table above, but the exact conversion isn't a simple 1:1 formula. Use the ADA's official concordance table if you need a precise old-scale-to-new-scale equivalent.

PAT 20 vs 22 Score Difference: Why Two Points Is a Big Deal

Here's the part that surprises people: the PAT 20 vs 22 score difference is only two raw points, but it's roughly a 10-percentile-point jump. At 20, you're beating about three out of four test takers. At 22, you're beating something closer to five out of six.

That gap sits right in the densest part of the bell curve, where most applicants cluster. It's exactly the range where competitive programs draw their cutoffs and where a two-point raw improvement genuinely changes how your application reads next to everyone else's.

Many dental schools publish average matriculant PAT and AA scores: pull those numbers directly from each school's admissions page or ADEA profile rather than trusting secondhand percentile claims. More competitive programs tend to cluster closer to the 22+ range on the old scale; others admit comfortably in the high teens to low 20s. Confirm current numbers with each school, since they shift class to class.

Why Chasing a Raw PAT Score Is the Wrong Goal

A "20" or "22" isn't a badge. It's only meaningful relative to other test takers and, more importantly, relative to what your target schools actually admit. Aiming for "a 22" with no context is the same mistake as aiming for "a good grade" without knowing the class average.

The fix: find the average matriculant PAT for schools you actually want, translate that raw number into a percentile using the chart above, then add a buffer of 5–10 percentile points so you're not landing exactly at the average of an already-competitive pool. That target percentile, not the raw score alone, is what you should practice toward.

This is the same philosophy behind everything we built at DATPractice: the DAT is a standardized test, so your consistent practice score is your real score. Practice correctly, track the percentile trend behind your numbers, and stop guessing at what a raw score "means."

What Percentile Should You Actually Target for Your Dream Schools

1. Pull the real numbers. Get each target school's published average matriculant PAT and AA, not a forum estimate.
2. Convert to percentile. Use the chart above (or the ADA's official table) to translate that raw average into a percentile.
3. Add a buffer. Target 5–10 percentile points above the average matriculant, since "average" means half scored lower and half higher.
4. Set that percentile as your goal. Not "get a 22." Get to the 85th percentile, and let the raw score follow.
5. Re-check periodically. Matriculant stats shift class to class, so revisit your target during your prep window.

Track Your PAT Percentile Trend, Not a Single Lucky Score

A single practice PAT score is noisy. One easy pattern-folding section or one unlucky cube-counting section can swing your raw score by a couple of points without reflecting your actual skill, and as you now know, a couple of points can swing your percentile a lot.

How to Actually Move Your PAT Percentile Up

Once you know your target percentile, the way to close the gap is subtype by subtype, not "more generic PAT practice":

• Keyholes: lock onto one orientation cue per shape instead of matching the whole silhouette: see our keyhole PAT tips guide.
• Top-front-end: anchor on one view and eliminate wrong answers from there, covered in our TFE method breakdown.
• Angle ranking: use pairwise elimination instead of estimating exact degrees, detailed in our angle ranking guide.
• Hole punching: track the punch through each fold with a consistent line-counting method: our hole punching strategy covers it.
• Cube counting and pattern folding: pick one fixed system for each and never switch mid-section.

Every miss on a full-length test points to a specific fixable gap in one of those six subtypes. Our AI tutor finds the concept behind each miss from your personal history, re-teaches it to test-depth, then builds unlimited custom practice tests from your own weak spots until the percentile trend moves.

PAT Percentile vs. AA Percentile: Don't Confuse Them

The PAT and AA are graded on completely different content and reported separately, so a strong PAT percentile tells you nothing about your AA percentile, and vice versa. Some schools weight PAT more heavily, since it's the closest proxy the exam has for manual dexterity and 3D spatial judgment tied to clinical work.

Want to see how your raw science scores translate into an AA? Our DAT score calculator walks through that conversion the same way this article walks through PAT percentiles.

Why Your Percentile Trend Is Only as Good as Your Question Supply

Everything above this section assumes your practice percentile actually reflects your PAT skill. It won't, for long, if you're pulling from a fixed set of questions. The PAT isn't knowledge you memorize once and keep. It's a trainable visual skill, and skills improve through reps: questions your brain hasn't already seen the shape of.

Here's the trap almost every PAT student falls into without noticing. Every practice resource out there, ours included by default, is built on a set of items: a few hundred keyholes, a few hundred cube stacks, maybe a couple thousand items across all six subtests. Work through a set like that and something quietly changes: you stop solving new spatial-reasoning problems and start recognizing individual ones. Your practice percentile keeps climbing, your confidence climbs right along with it, and neither one reflects the skill the real PAT is actually going to test. That's the worst combination you can walk into test day with: a number that looks great in your log and falls apart the moment you're handed geometry you've never actually reasoned through.

That's the specific problem we built our PAT generator to solve. Every item comes from a (seed, difficulty) pair, so the supply of new questions is infinite by construction. You won't see the same keyhole, cube stack, angle set, or folding pattern twice unless you deliberately re-run that exact seed. Each item is also independently checked: a separate solver re-derives the correct answer straight from the underlying geometry, and if it disagrees with the generator, the item never ships. Difficulty is a dial rather than a fixed setting too, so you can push your reps past what the real test throws at you instead of running out of hard questions right when you need them.

That's the difference between a percentile trend and an illusion of one. A trend built on a finite set plateaus quietly the moment you've seen everything in it, even as your confidence keeps rising. A trend built on a generator means the same thing on rep 100 that it meant on rep 1: how you actually perform against a question you've genuinely never seen before.