Why is my VO2max estimate dropping even though I'm training more?

Why "training more" doesn't guarantee a rising number

This is the part that trips people up: VO2max estimates respond to the quality and consistency of specific input data, not to total training volume. You can genuinely increase your weekly mileage or workout count while the number the watch reports goes down, because volume and the inputs the algorithm actually uses are different things. A runner who adds three extra easy treadmill sessions a week, for instance, has objectively trained more — but if none of those sessions are outdoor, GPS-tracked, and above the effort threshold the algorithm requires, they haven't added a single qualifying data point. Meanwhile, cardiovascular drift from accumulated fatigue (your heart rate creeping higher at the same pace, session after session, because you're under-recovered) is a training effect that genuinely pushes the estimate down — and it's also correlated with training more, which is exactly why the two get confused.

Algorithm-side causes vs. a real decline

Splitting these into two separate diagnostic paths is the fastest way to figure out which one you're dealing with.

Algorithm-side causes (the number moved, your aerobic capacity probably didn't):

• GPS or pace-data quality dropped. Tree cover, tall buildings, tunnels, treadmill runs, or a route with more stop-and-go segments (traffic lights, crowded trails) all feed noisier pace data into the heart-rate model.
• Your max-heart-rate estimate drifted. Both Apple's and Garmin's algorithms lean on your heart-rate reserve (the gap between resting and max). If the watch quietly revised its estimate of your max HR — common after a season without any near-maximal efforts — every subsequent calculation shifts with it.
• You stopped feeding it qualifying sessions. Apple Watch only calculates VO2max from Outdoor Walk, Outdoor Run, or Hiking workouts; Garmin's models are similarly GPS-dependent. A block of indoor training, cross-training, or lower-intensity work that never crosses the required effort threshold means the estimate goes stale rather than updating — and a stale number sitting on your screen can look like a drop even when it's actually just old.
• A software update changed the algorithm version. Firmware and OS updates occasionally retrain or recalibrate the underlying model. When that happens, the shift usually appears as a one-time step change that lines up with an update date, then stays flat — as opposed to a decline that keeps sliding week after week.
• You switched watch, wrist, or strap fit. A different device, a looser optical sensor, or a new chest strap can each introduce a small systematic offset in the heart-rate readings the estimate is built from.

Real-decline causes (the number moved because something in your physiology or training actually changed):

• Accumulated fatigue without enough recovery. This is the one genuinely tied to "training more." Piling on volume or intensity faster than you recover raises your heart rate at any given pace (cardiovascular drift), which the algorithm correctly reads as reduced efficiency — even though it isn't a drop in your ceiling capacity, just in how well your body is expressing it right now.
• Intensity dropped even though volume rose. VO2max responds specifically to sessions that push heart rate meaningfully above resting — easy-paced added mileage doesn't provide the stimulus that raises it, and can even mask a plateau by making you feel like you're doing more.
• Illness, heat, dehydration, or heart-rate-affecting medication. Any of these shift the heart-rate-to-pace relationship the estimate depends on, usually for one to a few weeks rather than permanently.
• A genuine, sustained loss of aerobic fitness. After several weeks off training, a major life-stress period, or a long illness, the decline can be real — the way to tell is that it persists across many sessions and often comes with a subjective sense of reduced capacity, not just a lower number on a screen.

Decision table: is it the algorithm or is it real?

A 2-4 week protocol for figuring out which one you're looking at

1. Log at least 3 qualifying outdoor GPS sessions per week at a clear, sustained effort — this rules out the single biggest algorithm-side cause (missing or low-quality inputs) before you conclude anything else.
2. Keep the route and conditions as consistent as you can — the same loop, similar time of day, similar weather — so you're not introducing new GPS or heat variables mid-diagnosis.
3. Track resting heart rate and how recovered you feel alongside the number. If VO2max falls but resting heart rate and subjective recovery stay normal, lean toward an algorithm-side cause. If resting heart rate is also elevated and you feel run-down, that points toward accumulated fatigue.
4. Check whether an update landed around the time the number changed. A dated, one-time step change is a strong signal for an algorithm-version shift rather than a real trend.
5. Look at the trend line over the full window, not the daily number. VO2max genuinely doesn't move much week to week — judge it on a rolling multi-week average rather than the most recent single reading, the same approach that applies to accuracy questions generally (see how Apple Watch's estimate compares to lab testing).
6. If it's still trending down after 3-4 weeks of consistent qualifying data, treat it as a real signal worth acting on — dial back volume or intensity for a short block, prioritize sleep, and reassess rather than pushing harder into a fatigue hole.

When to take it seriously

Most VO2max wobbles resolve on their own once you feed the algorithm consistent, qualifying data again. Two patterns are worth paying closer attention to: a decline that persists for a month or more despite normal training and recovery, and any drop that comes paired with unusual breathlessness or fatigue at efforts that used to feel easy. The second pattern in particular is worth mentioning to a doctor — a wearable's VO2max estimate is a fitness trend tool, not a diagnostic one, and it has no way to distinguish "you're undertrained right now" from a change worth medical attention.

Where this fits into the bigger picture

VO2max is one of the inputs Body Age weighs alongside resting heart rate, HRV, sleep, and step data, and Vita reads it as a trend from Apple Health rather than reacting to any single update — which matters given how much the number can move on data-quality grounds alone. If you're trying to figure out whether a specific stretch of low readings reflects your training load or just noisy GPS days, Vita's AI coach can look at your recent resting heart rate, HRV, and training pattern alongside the VO2max trend and help separate the two explanations instead of leaving you staring at one number. And if the honest answer turns out to be "not enough high-effort stimulus," a structured Zone 2 base with weekly harder intervals is the best-evidenced way to move the number in a direction you can trust.