This is a drill, not a lesson. You get 21 real-format coded METARs — the raw strings as they come off the wire, not a decoded summary — and questions that ask you to pull a specific figure out of one and do something with it. Ceiling. Visibility. Wind. Temperature and dewpoint spread. The same things the knowledge test asks, in the same unfriendly format.
It is built for the candidate who has already read a METAR explainer and nodded along, then froze the first time one appeared in a practice question with a clock running. Reading a report slowly at your desk and reading one under time pressure are different skills. If you still need the teaching, start with how to read a METAR for drone pilots and come back here to build the speed.
Recognition, mostly. A METAR is a fixed-order string, and once the order is automatic you stop reading it left to right and start jumping straight to the group you need. That jump is the whole skill. A question that asks for the ceiling does not want you to decode the altimeter setting, the remarks, or the present weather — it wants one number out of one group, and every second you spend on the rest is a second you did not spend on a regulations question.
The second thing it teaches is the step after the decode. Most exam METAR questions are two-part: read the report, then apply a rule to what you read. The rule is almost always 14 CFR 107.51 — 3 statute miles of flight visibility from the control station, 500 feet below a cloud, 2,000 feet horizontally from one. Candidates who lose these questions usually decode the report correctly and then subtract in the wrong direction. Drilling both halves together is the point.
How it mirrors exam conditions
Three things are deliberate. The reports are presented coded, because the real test never hands you a plain-English translation. The questions ask what the FAA asks rather than what is interesting about the report. And the answer options are shuffled on every attempt, so you cannot learn a position instead of an answer. The shuffle is fixed for a given attempt, so pausing and resuming does not move the options under you.
What it does not do is put a countdown on the screen. That is what the mock exam is for — 60 questions in 120 minutes, 70% to pass, answers hidden until you submit, and part of the Complete plan. Use this trainer to get the decode fast, then prove it under the real clock.
What the 21 reports cover
The reports are ordinary coded observations, so between them they carry the groups a Part 107 question is most likely to be built around. These are the ones worth being fast at.
| The group in the report | What the exam does with it | Where candidates lose it |
|---|
| A sky condition group | Finding the ceiling — the lowest BKN or OVC layer | Treating a FEW or SCT layer as a ceiling |
| A cloud height in hundreds of feet | Converting to feet AGL, then applying 107.51 | Reading BKN012 as 12 feet, or as MSL |
| Visibility in statute miles | Whether 107.51's 3 SM is met | Confusing reported visibility with flight visibility from your control station |
| A wind group with a gust | Recognizing direction, speed, gust factor | Assuming a wind limit exists in Part 107 — none does |
| Temperature and dewpoint | Spread, and what a narrow one predicts | Missing the minus sign on M03 |
VV002 in place of a layer | Indefinite ceiling, vertical visibility 200 feet | Skipping the group entirely |
The Z time group | That METARs are always UTC | Reading 041853Z as local time |
Wind direction in a METAR is referenced to true north, which is a detail candidates trip over because the tower reports magnetic. Neither figure changes whether you may launch, but the exam has asked which reference applies.
A worked example
METAR KDSM 041853Z 24018G28KT 4SM BR BKN007 OVC015 18/16 A2988 RMK AO2
You want to fly at 300 feet AGL on this field. Work it in the order that answers the question fastest.
Visibility first: 4SM clears the 3 statute mile requirement in 107.51, so that gate is open. Now the ceiling. BKN007 is the lowest broken layer at 700 feet AGL — heights in a sky condition group are in hundreds of feet above the reporting station, so 007 is 700, not 7. Section 107.51 wants 500 feet of clearance below a cloud, which caps you at 200 feet AGL. Your planned 300 feet is illegal. Descend, or wait.
Two things in that report do not affect legality but do affect the flight. The wind is 240 at 18 gusting 28 knots — Part 107 sets no wind limit, only an 87-knot groundspeed limit on the aircraft, so this is a manufacturer and pilot judgment call rather than a regulatory one. And 18/16 is a two-degree temperature-dewpoint spread with BR already reported, which tells you the visibility is more likely to fall than to improve.
Common mistakes
Decoding everything. You do not need the altimeter setting to answer a cloud-clearance question. Skim to the group, answer, move on.
Forgetting the report is AGL at the station. Sky condition heights are above ground level at the reporting airport. Your launch site may not be at that elevation, and the exam has used that gap.
Reading the reported visibility as your visibility. Section 107.51(c) measures flight visibility from the control station. An airport eight miles away reporting 6 SM is evidence, not proof. Your preflight assessment under 14 CFR 107.49 requires you to consider local weather conditions yourself.
Learning 21 reports instead of the format. If you find yourself remembering that the third sample was the one with the low ceiling, reset and work them cold.
Weather is 5% of the Unmanned Aircraft General knowledge test under the blueprint effective 29 September 2025, which is roughly three questions. That is a small slice, and it should get a small slice of your study time — the payoff here is that METAR questions are among the most mechanical points on the test once the format is automatic. The Airman Certification Standards sets out the knowledge areas behind them.
Pair this with the sectional chart trainer for the other figure-reading skill, and with the full practice bank when you want METAR questions mixed in with everything else rather than isolated. That bank holds 1,261 questions, 100 of them tied to a METAR, and it is part of the Complete plan. For the rule you are applying at the end of every decode, VFR weather minimums covers why the Part 107 figures are fixed while the Part 91 table moves.
Frequently asked questions
METAR practice means drilling coded weather reports until you can extract a figure without decoding the whole string. Part 107 knowledge test questions present METARs in raw coded form and ask you to apply a rule — usually the 3 statute mile visibility and 500-foot cloud clearance in 14 CFR 107.51. Speed matters as much as accuracy.
Weather is 5% of the Unmanned Aircraft General knowledge test under the blueprint effective 29 September 2025, which works out to about three of the 60 scored questions. Not all three will be METARs. However, weather figures also appear inside operating-limitation questions in other knowledge areas, so the decoding skill is worth more than the 5% suggests.
The trainer uses 21 METAR samples in real coded format, each with decoding questions attached. You work the raw string rather than a plain-English translation, which is how the Part 107 knowledge test presents a weather report. Answer options are shuffled on every attempt, so you cannot learn an option's position instead of the answer itself.
Work coded reports cold, pulling one figure at a time, and always finish by applying the rule rather than stopping at the decode. That rule is usually the 3 statute mile visibility and 500-foot cloud clearance in 14 CFR 107.51. Once the decode is automatic, take METAR questions inside a timed sitting rather than in isolation.
A decoder translates a report for you, which is exactly the crutch the knowledge test removes. This trainer shows the coded string and asks the question, so the translation happens in your head. It also asks the follow-up the exam asks — whether a proposed altitude is legal — rather than stopping at what the groups mean.