The Part 107 initial test is a multiple-choice knowledge test, not a flight check. Difficulty comes from the number of rules that sound alike and from charts, not from trick math. The current format is public. PSI's UAG bulletin, effective September 29, 2025, is the document this course uses for the weights. The ACS is the document that lists the knowledge areas. An older percentage table still printed in the ACS is obsolete for the scored test. Use the bulletin.
The format
You get 60 scored questions and 5 unscored validation questions. You have 120 minutes. The passing score is 70 percent on the scored items, which is 42 of the 60. You will not know which five are unscored, so answer all of them. The ACS says only one choice is complete and correct. The others are incomplete or wrong. If a question is hard, mark it and come back. That is the ACS's own tip, and the testing software supports it.
The FAA publishes Airman Knowledge Test statistics, including volume and pass rates for Unmanned Aircraft General – Small. Those figures are on the FAA test-statistics page, and they change by year. A sample quiz on a vendor site is not that report, and it is not the test. PSI's own materials say a free sample does not represent the full range or difficulty of the live exam. Do not treat a round number from a training advertisement as the FAA statistic.
Where the points are
Regulations are 48 percent. That is almost half the scored form: who may fly, registration, remote pilot duties, the operating limits, Remote ID, night, operations over people, waivers, and reporting. If you only drill charts, you are studying the smaller slice.
Operations are 25 percent. Human factors sit inside operations, not in a sixth weighted area. Radio phraseology, airport patterns, decision-making, and physiology live here.
Airspace is 20 percent. Classes, chart depictions, authorizations, and special-use airspace are in that fifth of the test. Weather is 5 percent. Loading and performance is 2 percent. Those last two are small and still predictable. A METAR decode or a Figure 2 load-factor multiply is worth learning because the items are short once you know the code. They are a poor place to spend all of your time.
What the hard questions feel like
They stack two rules. A cloud deck at 1,200 feet AGL, with no structure nearby, looks like a cloud-clearance problem: 500 feet below the base would suggest 700 feet AGL. The 400-foot cap in §107.51(b) is lower, so 400 feet AGL is the answer. People who remember only the cloud numbers miss it. The sample question on this page is that item.
Chart questions feel hard when the figure is busy. The skill on those items is locating and interpreting airspace, airport, obstruction, and frequency data. It is not reciting the legend from memory with the figure closed, and it is not memorizing every airport. When the stem says "refer to Figure 22," open Figure 22. Guessing without the figure is how those items are missed. From October 26, 2026, some images will be current sectionals that are not in the supplement. The reading skill is the same.
How to prepare without boiling the ocean
Study regulations first, then operations, then airspace, then a focused pass on weather codes and the load-factor table. Use the free 20-question practice test as a weighted sample, not as a score that predicts the PSI result. It is 20 items. The UAG scored section is 60. Then use full-length practice under a timer.
Test-day mechanics, the fee, and what you carry to the center are on the test day page. The airspace weight is why the airspace class page is worth more time than density altitude, even though both can show up.