Start with the thing almost every swimming website gets wrong: freestyle is not a stroke.

“In a Freestyle race, Freestyle means an Athlete may swim any style.” — World Aquatics Competition Rules, Part Two, Article 5.1

That is the entire rule. You may swim a freestyle race doing breaststroke, backstroke, doggy paddle, or anything you invent on the day. Front crawl is simply the fastest thing anyone has found, so that is what everybody chooses. The event is a category; the stroke is front crawl.

This matters more than pedantry, because it tells you what the rulebook actually cares about — and it’s far less than people assume. Below: what the rules require, what research has measured, and what I see at the poolside. Kept separate, so you can tell which is which.

What the rules actually require

Quoted from the World Aquatics Competition Rules, version 18 February 2026, Part Two. There are only four freestyle rules, and none of them mention technique.

Touch the wall every length.

“At least one part of the Athlete’s body must touch the wall upon completion of each length and at the finish.” — Article 5.2

Any part, and only one. This is why the tumble turn is legal in front crawl but not in breaststroke or butterfly, where both hands must touch simultaneously. Your feet hitting the wall in a tumble turn satisfy the rule.

Break the surface, except for 15 metres.

“At least one part of the Athlete’s body must break the surface of the water throughout the race, except that the Athlete may be completely submerged during the turn and for a distance of not more than 15 metres after the start and after each turn. By the 15 metre-mark, the Athlete’s head must have broken the surface of the water.” — Article 5.3

A rule that changed in 2026.

“Once some part of the head of the Athlete has passed the 5 metre-mark, immediately prior to reaching for the finish wall, the Athlete may be completely submerged.” — Article 5.4

The 2023–2025 rulebook had no Article 5.4 — freestyle had only three rules. The current edition adds an explicit allowance to submerge inside the final 5 metres. If you learned otherwise, that’s now out of date. The equivalent change was made to backstroke at the same time.

And on the start:

“On the Starter’s command ‘take your marks’, the Athletes will immediately take up a starting position with at least one foot in contact with the front of the starting platforms. The position of the Athlete’s hands is not relevant.” — Article 4.1.3

No rule about your hand entry, your body roll, your kick, your breathing side, or how high your elbow is. Every one of those is coaching judgment. Useful judgment — but not law, and worth knowing the difference when someone tells you your child is doing it “wrong”.

Where the propulsion actually comes from

Here is where honest sourcing gets interesting, because the research does not give one clean number.

A 2025 study in Sports states the established position: “87–90% of crawl swimming propulsion is generated by the action of the upper limbs.”

But a 2015 study in BioMed Research International measured it directly. Morouço and colleagues put 23 national-level young swimmers (12 male, 11 female, around 15 years old) through 30-second maximal fully tethered efforts — whole body, arms only, legs only — with a load cell sampling force at 100 Hz. Their result:

  • Males: arms 70.3%, legs 29.7%
  • Females: arms 66.6%, legs 33.4%

So: two-thirds, or nine-tenths? Both, depending on how you measure. Tethered swimming holds you in place, which changes the mechanics — the authors say so themselves, cautioning that tethered conditions “may not fully represent free swimming mechanics” and that their findings reflect short, high-intensity efforts. It is not a disagreement about swimming so much as a disagreement about measurement.

What survives either way: your arms do most of the work, and it isn’t close. Anyone who tells you front crawl is mostly legs is wrong under every method anyone has used.

But — and this is where coaching diverges from the percentage: the legs’ job in front crawl is mostly not propulsion. It is holding your hips up. A swimmer with a dead kick drags their legs low, presents a much larger surface to the water, and has to fight drag their whole life. Research on tethered force will score that kick as contributing little propulsion, and still your stroke falls apart without it. Low percentage, high importance. Those are not contradictory.

For a beginner, this ranks your problems: body position first, arms second, kick third — and kick as a body-position tool, not an engine.

The hand does not simply push backwards

The same 2025 Sports paper models the hand as generating both drag and lift, with the angle of attack — how the hand is tilted relative to its direction of travel — observed between 20 and 70 degrees. It also estimates the forearm adds roughly 8.5% to the effective propulsive area beyond the hand alone.

Two practical consequences. Your hand is a hydrofoil, not a paddle, so its angle matters as much as its size. And your forearm is part of the paddle — which is the actual justification for the “high elbow catch” cue every coach repeats. Dropping the elbow takes the forearm out of the equation.

I’d note this is an area of live research rather than settled fact, and I’d be sceptical of anyone describing the ideal hand path with great confidence.

Breathing: what the evidence supports

Bilateral breathing — every three strokes, alternating sides — gets recommended universally. There is decent evidence for it, and one result that surprises people.

Barden and Barber (2022, Sensors) studied 18 elite competitive swimmers (11 male, 7 female, around 19 years old) at submaximal speed, comparing unilateral breathing, bilateral breathing, and a snorkel. They found peak hip roll to the preferred side was significantly greater when breathing unilaterally — that is, always breathing to one side rolls you further that way, exactly as coaches suspect.

Bilateral breathing improved that symmetry. But the snorkel improved it more. Their conclusion is blunt:

“The results demonstrate that the most effective way to balance hip roll kinematics and mitigate roll asymmetry is to use a swim snorkel.”

Caveats, honestly: 18 elite swimmers, at submaximal speed, measuring hip roll — not race performance. It does not prove bilateral breathing makes you faster, and it certainly doesn’t mean a beginner should be forced onto their weak side before they can breathe at all on their strong one.

What I take from it: if a swimmer has an entrenched one-sided roll, a snorkel is a better corrective tool than nagging them to breathe bilaterally. That’s a genuinely useful thing to know, and it’s the opposite of the standard advice.

The faults I actually see

Coaching judgment, not research. In the order they cost you most.

Low hips. The root of nearly everything. Usually from looking forward instead of down, or a dead kick. Everything else you fix is wasted while the legs are trailing.

Lifting the head to breathe. Turn the head, don’t lift it. One goggle in, one goggle out. Lifting pushes the hips down and stops the stroke dead.

Dropped elbow in the catch. The arm sweeps back with a straight, low elbow, which takes the forearm out of the pull entirely — see the 8.5% above. This is the single biggest difference between someone who looks fast and someone who is fast.

Crossing the centre line on entry. The hand enters past the midline, the body snakes, and the swimmer fights their own stroke.

Kicking from the knees. Big bent-knee cycling, lots of splash, no propulsion and considerable drag. The kick starts at the hip.

Holding the breath. Extremely common in adults. You must exhale continuously while your face is in the water — if you hold it, you arrive at the breath needing to both blow out and suck in inside a fraction of a second, and you’ll never keep up.

Front crawl and SwimSafer

Worth knowing: SwimSafer Stage 2 asks for 25 m on the front in any stroke, plus 15 m on the back. Any stroke — so front crawl is not strictly required to pass it, and plenty of Singapore children clear Stage 2 on breaststroke.

There’s a pleasing symmetry with Article 5.1 here: at both the beginner and the elite end, the requirement is a distance, not a technique.

That said, front crawl is what most children want, because it’s what looks like real swimming to an eight-year-old. That motivation is worth using.

Practising it in a Singapore public pool

Lane etiquette first: keep right, overtake only at the wall, and if you’re doing drills, use the slow lane. A 25 m public lane at Bishan or Hougang on a weekend is not the place for repeated 25 m sprints.

Fix body position before anything else. Kick on your side with one arm extended, face down, rolling to breathe. If your hips sink, no arm drill will help you yet.

For breathing, exhale into the water like a slow sigh, continuously, every stroke. Most adult front crawl problems are breathing problems, and most breathing problems are exhale problems.

And a blunt one: doing more lengths badly makes you better at swimming badly. Front crawl punishes repetition of a bad pattern more than any other stroke, because you do so many cycles of it.

Where to start

If your front crawl has never felt right — you’re out of breath by 25 m, or you work hard and go nowhere — that’s a body position and breathing problem in almost every case, and it’s what our stroke correction work exists for. It’s popular with triathletes for exactly this reason.

Your first lesson is from $30+ for 45 minutes, with no commitment.

WhatsApp 9677 3858 — tell me what happens when you try to swim 50 m, and I’ll tell you what’s actually going wrong.

Sources