Applicants ask what score passes. The more useful question is what score gets called, because most locals rank everyone who qualifies and fill classes from the top down. Here is how the scoring and ranking work, what the published minimums really say, and why a cutoff number means nothing until you know which test your local gives.
This is the part almost every “what score do I need” answer online skips. The UA does not run one national entrance exam. Committees choose their own testing provider, and the three UA policies below — all published, all linked at the foot of this page — use three different tests with three incomparable minimums:
| Local | Test used | Published minimum |
|---|---|---|
| Local 447 — Sacramento, CA | Its own exam: about 3 hours, mathematics through first-year algebra | 65% to qualify |
| Local 7 — Latham, NY | Wonderlic Basic Skills | 50 or higher to be interviewed |
| Local 773 — Queensbury, NY | WorkKeys assessments | Average of at least 4.334 out of 7 |
The widely repeated figure for the GAN battery is around 70% to qualify, with some pipefitter and steamfitter locals setting it higher. Treat that as a planning assumption rather than a fact about your local: it is not published by GAN, and a committee is free to set its bar wherever its applicant pool and seat count put it.
The mechanism is consistent even though the numbers are not. GAN (or whichever provider) scores the battery and reports to the training committee, which applies its own standard in one or two stages:
Classes then fill from the top of that list as contractor demand allows. This is why “I passed and never heard anything” is such a common story and is not usually a mistake by anyone: a list is an ordered queue, and a minimum-clearing score puts you at the back of it. Lists are commonly active for one to two years, and results are usually reported back in about 60 days — both worth confirming with your training office, because both are local.
Section weighting is set by each committee and is not published nationally, so anyone telling you mechanical comprehension is worth 15% is inventing it. What the battery itself tells you is where the questions are: computation, number series and problem solving are 73 of about 140 questions — so arithmetic dominates a raw total no matter how the sections are weighted. If you only have time to fix one thing, fix the arithmetic.
| Section | Questions | Share of battery |
|---|---|---|
| Reading Comprehension | 42 | 30% |
| Numerical Computation | 28 | 20% |
| Numerical Reasoning | 10 | 7% |
| Problem Solving | 35 | 25% |
| Paper Folding | 12 | 9% |
| Mechanical Comprehension | 13 | 9% |
Question share is not the same as score weight, and this table is the former. It is still the best guide available for where to spend study time, because a section you cannot finish costs you points in proportion to how many questions it holds.
As high as you can get, which sounds like a dodge and is actually the only defensible answer on a ranked system. There is no score at which you are safe, because the bar is set by the other applicants in your cycle and by how many apprentices the signatory contractors are funding that year. Neither is published in advance and neither is under your control.
What is under your control, in rough order of return:
Diagnose by section before you study anything. Then work in this order: worked practice questions to find the weakness, the section deep-dives for the technique, then full-length timed tests for the pacing. If a second attempt is on the table, read the retake rules first — some locals cap attempts, and one publishes a remediation requirement you have to satisfy before you may test again.
There is no national UA passing score, and a number quoted without naming the test it belongs to is not an answer. Published UA criteria run on different instruments and different scales: 65% to qualify at Local 447 in Sacramento on its own math exam, 50 or higher on the Wonderlic Basic Skills test at Local 7 in Latham, an average of at least 4.334 out of 7 on WorkKeys at Local 773 in Queensbury. Around 70% is widely repeated as a GAN qualifying figure, but GAN Human Resources publishes no cutoff itself, and the qualifying standard belongs to your training committee in any case.
Usually both, in that order. Most committees set a qualifying minimum to advance you to the interview, then combine your test result with a scored oral interview into a composite that orders every qualified applicant. Classes fill from the top of that list down as contractor demand allows, so clearing the minimum is necessary and rarely sufficient — a middling score can sit on an active list until it expires.
Higher than this cycle's applicants, which is a number nobody can publish in advance because it depends who else applied and how many seats the contractors fund. That is the honest answer, and it has a practical consequence: treat the minimum as a floor to clear early and then keep training, because there is no score at which you are safely done.
Weighting is set by each training committee and is not published nationally, so we cannot tell you. What can be said from the battery itself is that the math-heavy sections carry the most questions — computation, number series and problem solving together are 73 of about 140 — so arithmetic fluency has the largest effect on a raw total regardless of how the committee weights sections. Ask whether yours uses a composite or per-section minimums.
Ask your training center when you confirm the appointment. GAN does not publish its scoring method, and this is a question where a confident guess costs points either way: if there is no penalty, leaving a blank throws away a chance, and if there is one, blind guessing is worse than skipping. It is a fair and specific question to ask, and the answer applies to your test.
About 60 days is the figure usually reported, delivered through the training center rather than to you directly from the testing session, and eligibility lists commonly stay active one to two years before a new cycle begins. Both vary by local, so confirm them with the training office — and note that if your list expires before your number comes up, you generally reapply and retest rather than carry the old score forward.
They can where a committee awards them — for veterans or for graduates of a recognized pre-apprenticeship program, for instance — and because the selection procedure has to be written down and applied uniformly, any preference your local grants is documented rather than discretionary. Ask what your local awards and what documentation it wants, and bring the paperwork with the application rather than after it.
The published minimums above are linked so you can read them in the original. They are examples of what UA criteria look like, not a statement about your local — and local policies change without notice.
Ten full-length timed batteries, 1,500+ questions across all six sections, and Piper to explain anything you miss — built for the part of the process that decides your place in line.
Download on the App StoreiPhone & iPad · free to download · 7-day free trial on Pro