Swimming and Apple Watch: Why Your Heart Rate Looks Wrong in the Pool (and How to Fix It)

You finish a hard 1,500 metres, breathing like you just sprinted up a hill, and your Apple Watch reports an average heart rate of 118 bpm with a graph that looks like a flat line someone stepped on. Or worse, the trace has a hole in the middle of your main set, or a spike to 190 that you definitely did not feel.

Swimming is the one workout where almost everybody's wrist heart rate data looks suspicious, and it is not just your watch. It is the hardest common activity for any optical sensor to measure, and there are two separate reasons the numbers look off — one is a genuine measurement failure, and one is not a failure at all. Separating them is the difference between chasing a bug in your watch and reading your swim correctly.

YouTube: Related video

TL;DR: Wrist optical heart rate in water is genuinely unreliable. In a controlled comparison against a chest strap, wrist-based watches went from roughly 4-8 percent mean error on land to 17-30 percent in the pool, with a systematic tendency to under-read, especially at higher intensities. On top of that, heart rate during swimming is physiologically lower than during running at the same effort, so your land-based zones will always overstate how hard a swim felt. Apple Watch is water resistant and handles swimming fine as a device — Series 2 and later are rated for shallow water activities — but the heart rate it reports in the pool deserves scepticism. The fixes: snug strap above the wrist bone, use the Pool Swim workout so Water Lock engages, expect gaps, judge swim intensity by pace per 100m and RPE rather than average heart rate, and set separate swimming zones if you want zones at all.

How Apple Watch measures heart rate, and why water breaks it

Apple Watch uses optical photoplethysmography — green LEDs shining into the skin and photodiodes measuring how much light comes back. Blood volume changes with each heartbeat change the amount of light returned. It works well on a dry wrist in steady-state motion, which is why it is surprisingly decent on a treadmill and unreliable on a rowing machine.

Water attacks that setup in three specific ways.

Sensor-to-skin contact. The whole method depends on the sensor sitting flat against skin with nothing in between. A film of water, moving water, or a strap that shifts when you push off the wall all insert something between the LEDs and your blood. This is the big one, and the reason sensor placement matters so much in water.

Motion. Every arm recovery in front crawl rotates your wrist through a large arc. Optical sensors do not tolerate rapid changes in orientation and pressure well. The swim stroke is close to a worst case: high arm speed, changing hydrostatic pressure on the sensor, and a constantly changing distance between sensor and vessels.

Physiology. Cold water and hydrostatic pressure narrow peripheral vessels, which reduces the pulsatile signal the sensor is trying to detect. Less signal, more noise.

The research backs this up clearly. In a 2022 study in Sensors, researchers tested two wrist-worn smartwatches against a Polar H10 chest strap in 10 expert swimmers, on land and in water. Mean absolute percentage error rose from about 4 percent on land to roughly 30 percent in the water for one device, and from 8 percent to 17 percent for the other. During in-water activity, one watch showed a mean deviation of around 57 bpm and the other around 18 bpm, both in the direction of under-reading. Their conclusion was blunt: water and arm movement are serious interference inputs, and performance gets worse as intensity goes up.

A 2020 PLOS ONE study found something more useful for practical purposes. In front crawl swim training, an optical sensor worn at the temple (the Polar OH1) was valid, while the same technology worn on the wrist was much less accurate. Location matters more than brand. The wrist is the worst place to measure heart rate in water, and the temple or upper arm is a much better one — which is exactly where swim-specific devices put their sensors.

Part of this is not a bug: swimming heart rate really is lower

Before you blame the sensor for everything, know that swimming produces genuinely lower heart rates than land exercise at the same perceived effort, for reasons that have nothing to do with the watch.

• Water immersion shifts blood centrally, increases stroke volume, and triggers a mild bradycardia. Being horizontal, immersed to the chest, lowers heart rate at rest.
• Breath-holding and face immersion activate the diving reflex, which slows the heart further. You get a small dose of this every time you put your face in the water.
• Body position — prone and supported by water rather than upright against gravity — changes the whole cardiovascular picture.
• Cooling lowers heart rate when water is below body temperature.

The numbers: a 2019 study in Sports comparing maximal heart rates found swimmers typically reached a lower maximum in water than on land, with an average difference around 7 bpm versus running, though the spread between individuals was wide (about ±5-6 bpm). An older study of peak heart rates in maximal running and swimming suggested subtracting roughly 12 bpm from your treadmill maximum for young adult fitness swimmers.

The practical consequence is important and most people miss it. If you set your zones from a land-based max heart rate and then swim, your watch will put you in a lower zone than your actual effort, because both the sensor and the physiology are pushing your reported heart rate down. A swim that felt like Zone 4 shows up as Zone 2. Then you look at your weekly zone balance, conclude you are training too easy, and push harder than you should.

The failure modes and how to fix them

Gaps in the trace. The watch lost lock for 20-60 seconds. Common during push-offs and turns. Fix: tighten the strap by one notch before you get in, and wear it a finger's width above the wrist bone rather than over it.

Heart rate stuck low during hard sets. The signal is weak and the algorithm is reporting something conservative. Fix: there is no complete fix on the wrist. Accept it, and use a workout with pace or RPE-based targets instead of HR.

Spikes. Almost always artifacts. Do not panic and do not let them ruin your session — look at the shape of the curve, not the peak value.

Zones that look wrong. Set dedicated swimming zones or work from perceived effort (see below).

Water Lock confusion. On Apple Watch, Water Lock is what stops the touchscreen responding and ejects water through the speaker after your swim. With Apple Watch Series 2 and later, pool and open water swim workouts are supported. The important detail: start the workout before you get in, because that is what turns on Water Lock and puts the watch into its swim mode.

Wrong pool length. If your distance is off by a consistent amount, you probably entered 25 metres for a 25-yard pool, or the other way around. Check the pool length setting in the workout.

Apple Watch uses its own optical sensor in the pool. It is not going to record heart rate from a chest strap during a swim, because Bluetooth does not transmit reliably through water. If chest-strap accuracy during swimming is a priority for you, that is a Garmin-style setup where an HRM-Swim or HRM-Pro strap stores the data internally and you sync it after the session.

How to set your watch up for cleaner swim data

A checklist that actually changes the output:

1. Strap snug, one notch tighter than for running. Not so tight it restricts circulation, but no sliding.
2. Position above the wrist bone, on the flatter part of the forearm.
3. Start the Pool Swim workout before entering the water, and confirm Water Lock engaged.
4. Set the correct pool length and, if available, the correct stroke type — this drives lap and distance accuracy.
5. Expect missing heart rate and do not fight it. If a stretch is empty, it is empty.
6. Log your RPE after each session. One number, and it will be more useful than the average heart rate your watch reports in water.
7. If you want reliable in-water heart rate, move the sensor off the wrist — a temple or upper-arm optical sensor is the research-supported choice, or a swim-specific chest strap.

What to track instead of average heart rate

If the heart rate trace in the pool is unreliable, what do you actually use to judge swim training?

Pace per 100m. The most objective and comparable metric you have. If your pace holds steady across sets with the same rest, you are fitter than last month. If it fades, you are fatigued or you went out too hard.

Rest heart rate and recovery between intervals. Stop, stand, and look at your watch 30 to 60 seconds after an interval. The rate at which it drops tells you about your fitness, and it is measured while you are relatively still, so it is more trustworthy than mid-stroke readings.

Stroke count and distance per stroke. Efficiency data that no optical sensor can ruin.

RPE or the talk test. In the pool, because heart rate under-reports, perceived effort is genuinely the better guide. A swim set that leaves you able to speak only in short bursts is a hard set, whatever the trace says.

Distance and duration total. If you want one honest load number from a swim, distance times effort is more reliable than heart rate time in zone.

Next-morning recovery metrics. This is where swimming pays off in your data honestly. An overnight heart rate that is 4-6 bpm above baseline and HRV below baseline the morning after a hard swim session tells you the session was real, even if the in-pool trace looked flat. That is the signal Century uses to adjust your daily scores.

Setting swimming zones that make sense

You have two decent options.

Option 1 — the arithmetic approach. Take the maximum heart rate your watch estimates for running, and subtract roughly 10-12 bpm for swimming. Then set swimming zones from that number. It is a crude correction, and the individual spread is wide enough that it will be wrong for some people, but it is far better than using land zones unchanged.

Option 2 — test in the water. Warm up, then swim a hard 200-400m effort with good rest before it, and take the highest number you see. Do it twice on separate days and use the higher one. It is a better estimate than any formula, subject to the sensor caveats above. I go through the setup mechanics in Apple Watch heart rate zones.

If neither appeals, skip zones for swimming entirely and train by pace and RPE. A swim-specific zone setup built on noisy data can be worse than no zones at all.

Does any of this matter for your health data?

It matters less than you think for health, and more than you think for training decisions.

For health tracking, the useful swim signals are not in-pool heart rate: they are that you did the session at all, how long, how hard it felt, and how your overnight recovery responded. Aerobic training effects accumulate from the work, not from a clean heart rate curve.

For training, the risk is specific and real. If you trust a heart rate trace that under-reads, you will conclude your easy swims were too easy and your hard swims were mediocre. You will then overreach on the easy days and under-do the hard ones, which is the classic mistake of training from bad data. Trust pace, effort and next-morning recovery, and treat swim heart rate as a nice-to-have that is sometimes approximately right.

Bottom line

Apple Watch is a genuinely good swimming watch for everything except wrist heart rate. It counts laps, logs strokes, tracks distance and pace, and it is rated for shallow water use. The heart rate number, though, is fighting water, motion and physiology at once, and controlled testing shows wrist optical sensors can be off by tens of beats per minute in the pool, with a bias toward under-reading.

So do this: fix the strap and the sensor position, start the workout before you get in, expect gaps, and judge swim intensity by pace per 100m and how hard it actually felt. Set swimming-specific zones or drop zones for swimming altogether. Then check your overnight heart rate and HRV the next morning — that is the part of your swim data your watch gets right, and it is the part that should shape what you do the next day.

Century AI helps you understand your body with a daily health score, recovery score, and sleep insights — using the watch you already wear.