Ask whether a personal trainer is worth it and the internet splits in two. Gyms and trainers say yes, forums say keep your money and read up, and none of the top pages we read pointed to a single trial.
Three of the trials we examined ran the obvious test, the same plan coached and uncoached. The clearest edge showed up in exactly the people the sales pages say may not need a trainer.
In the newest of those trials, from 2025, people who had lifted for at least a year followed the same 10-week plan. Some had a coach in the room every session; the rest used a coaching app or a written plan. Among those who finished, the coached lifters added about 7 kg more to their squat than the rest.
That's about a quarter of a kilo of extra squat per planned session. The coached group was small, though, only 15 finishers, younger than the others and with the lightest squats at the start, which is why two older trials matter.
They agree on the direction. A trial from 2000 found the same squat edge in men with a year or two of lifting behind them, and one from 2023 found a smaller edge that could be close to nothing. The bench press was another story: two of the three trials found coached and uncoached lifters gaining about the same there.
Here's what the sales pages leave out: every one of those trials studied people who already lift. Beginners weren't in them, so the trials can't say what a coach does for a beginner's squat, either way.
None of the seven studies measured cost, so none can tell you what a session is worth. They can show what the coach actually did. That makes the job, not the price, the fair way to compare paid help, which is how a trainer's squat edge sits next to a meal box's month-six slide.
It wasn't just counting reps. In the 2000 trial, the coached lifters raised their training weights faster, week after week. In the newest one, the written plan spelled out when to add weight, but the coach adjusted the load and fixed technique as the lifters trained.
Two smaller clues point the same way. Women used to training with a trainer picked heavier weights even when tested without their trainer, though that's a link, not proof. And in classes of untrained men, one coach for every 5 people beat one for every 25 on strength.
But the sceptics get one thing right, and it may be what you were hoping to buy.
The mirror barely noticed
A lot of trainer marketing sells a body, a new shape in 30 days with a before-and-after to prove it. A 2022 review led by Fisher combined the six of its 12 trials that measured body composition. People with and without a coach ended up changing about the same.
And the single trials here weren't 30-day sprints: they ran 8 to 12 weeks. In the newest one, where everyone was asked to keep eating as usual, no group lost a measurable amount of fat.
If weight loss is the real goal, these trials can't speak to it: none of them put anyone on a calorie cut. And whether a meal plan can do that job is a different question.
With a coach, three smaller trials found some extra lean tissue, meaning muscle and everything else that isn't fat. Two measured it with an electrical body test whose precision the newest trial's own authors called limited, and in the third the coached group also got a different programme. For people who already lift, whether a coach changes the body is a fair question, not an answer.
The quarter you'd give up, and the catch
If a trainer isn't in the budget, this is the number to hold onto. Lifters who trained without a coach's help on the same plan, and made it to the end, kept about three-quarters of the squat strength the coached lifters gained.
In the newest trial, the written-plan finishers added 19.4 kg against the coached group's 26.6. The oldest trial landed in almost the same place, 25 years earlier. The coach's share was roughly the last quarter.
Not the whole thing.
That result only holds if you get to the end. In the 2025 trial, 15 of 17 coached lifters got through at least 26 of the 30 sessions. With the written plan, 25 of 47 did, about half, and nobody tested that gap statistically.
That's one trial, though. When earlier trials were combined, the gap shrank to a few points: people kept to close to 9 in 10 of their planned sessions with supervision, and a little fewer without it.
The honest version of your question may not be whether a trainer is worth it. It may be which of those groups you'd have landed in.
How alone is alone?
One trial's big gap and the combined trials' small one may have the same explanation: how alone the alone groups really were.
In the newest trial, the written-plan group was checked only when its logs came back at the end, and about half fell short. With an app that sent reminders, about 1 in 5 did. And in the 2000 trial, where the trainer stood in the room for both groups, everyone who stayed came to every session.
The combined trials sat in between: several of their unsupervised groups had some instruction or check-ins, and the gap was small.
Line those up and a pattern appears: the more alone people were, the more seem to have fallen short. No trial in our set tested check-ins on their own, or a training partner, so it's a hint rather than a finding.
It suggests the help that keeps people going may be someone keeping tabs, more than someone standing beside you. Online coaching might do that job.
The newest trial tested it, and the answer split in two.
The app nearly matched the coach in the room on finishing, about 8 in 10 against about 9 in 10. But its finishers added 19.2 kg to their squat, about the same as the written plan, while the coached group added 26.6 kg.
Every trial in our set that found the squat edge had a coach physically there, giving feedback as each lift happened. The app's feedback on filmed lifts came later, and the trial didn't test whether that timing is why.
If the app you're weighing is a weight-loss app with a coach added, that's a different job, judged on the scale, and the trials on it haven't shown a clear difference or ruled one out.
The squat, the mirror or the finish line
Where all of it lands depends on what you want.
If you already lift and want a bigger squat, the evidence points to a real edge from a coach in the room, on the same plan you'd follow anyway. If you're paying mainly for a different-looking body, it points to almost no difference from the coaching itself, though that question isn't closed.
If you'd finish a good plan on your own, the research suggests you'd keep most of the strength without one. If you've started and stopped before, a coach, or maybe just someone checking in, may matter most. The newest trial's authors read it the same way: a coach matters most for complex lifts and for people who need someone to answer to.
None of these trials tested the in-between, a short block of coached sessions to learn the lifts before going it alone. And since no study measured cost, the price is yours to weigh against all of this.
Who wasn't in the room
Nobody over 58 took part in any of the single trials, and the trials of people who already lift ran 8 to 12 weeks, so what happens over a year is beyond them.
Beginners, the people trainers are sold to most, get only looser evidence. It points the same general way: a small strength edge with supervision in combined trials of mostly untrained people, and bigger strength gains in classes with five coaches instead of one. Whether a trainer keeps you from getting hurt, our evidence simply doesn't answer.
None of the seven studies we examined compared a group class with a personal trainer. If your real goal is weight loss, the question changes shape. The weight-loss trials we gathered tested programmes, not personal trainers, and judged them on the scale.
There, the premium-looking option came out behind: across seven trials, people in group programmes lost 1.9 kg more at a year than people with one-to-one support.
There's a catch: the groups also got more hours of contact where the trials counted them, and the trials disagreed a lot on the size. Whether it was the company or the hours, the review's authors couldn't tell.
One trial put three ways of following the same plan side by side. A coach in the room bought the extra squat strength. A coaching app with feedback on filmed lifts kept almost as many people going, but its finishers' squats grew about as much as the written-plan group's.
The written plan, checked only at the end, got about half its lifters to the finish, and those who made it still kept most of the strength. Nobody tested those finishing gaps, and when earlier trials were combined, sticking to the planned sessions differed by only a few points.