Garmin VO2 calculation

Hey guys,

My Garmin connect app makes a VO2 calculation higher than my watch and then the watch vo2 ends up overwriting it. I connect the watch to hrm and power meter and iPad to power meter, trainer and hrm. I discard the ride on watch in order to not have duplicate rides but still get the training stress on the watch but it calculates it lower. Today the tr app and Garmin connect gave me 69 which then got overwritten by watch as 66?

I realise the Garmin vo2 isn’t the end of the world or probably v accurate but I like to track it. I have a Garmin fenix 6 pro for reference.

Any one else finding this?



1 Like

It’s well within the error margins of those VO2 estimates. I guess the watch has more data to go on and is the more accurate as a result?

One of the Running channels on YouTube did a test and found it to be very accurate considering what it has to work with.

1 Like

:open_mouth: I know its only a n of 2 but I didn’t expect it to be that close.

I’ll do a full performance analysis later this year and finally get to compare how close the two figures are for me.
I hope it won’t bump down the number too much :sweat_smile::


I get similar results from doing TR workouts using the PC app and doing the workout in Zwift. TR will get one VO2, and Zwift with get another. They’re about the same each time, but I just keep deleting the TR rides to, like you, not have the double training load in Garmin. Also, I notice the TR VO2 is just general VO2 vs Zwift coming up as cycling VO2.

I was up around 70-72 at my peak before my A-event. Back down to 60-62 nowadays without doing a training plan. I want to get a lab test one day (when I’m reading high again haha) to see how far off it is. It seems high considering my lower FTP, but who knows.

Lol, it dragged me into the youtube click phenomena. I quite like this video which says similar and explains how it can be skewed but not significantly :+1:

VO2MAX ON YOUR GARMIN: How do watches estimate VO2max?! - YouTube

Some things Garmin does with their hardware is really great, this is an example. Other things like ClimbPro, still need work.


The video I’ve been watching thinks it can be out by 5 but thats still phenomenal (I’ve never been above 75 on the Garmin prediction) :clap:

1 Like

I have an actual gas exchange reading from May 2018 and assuming my fitness it pretty close to then my vo2 max estimate from HRV4Training is pretty spot on.

I tend to see the number as a relative number as it relates to health outcomes. It serves me no purpose in shaping my training.

1 Like

Yip I’m skeptical of mine; it does show a general trend relating to health if you ignore the numbers. Just now I’m slightly above my readings 5 years ago when I got a power meter/ started structured training. It rose gradually as I started training then fell significantly in 2018 when I had cancer (alarmingly if you did pay attention to numbers it indicates that I was still healthy at 49 :open_mouth:; I prefer just to consider it a significant drop :nauseated_face: :face_vomiting:). As I got healthy and motivated I pushed the number up to a higher number before a gradual decline this year but as said still slightly above 5 years ago.

Another n=1 data point. I did many lab tests and consistently recorded in the 64-65mL/kg/min range.

My watch tracks very closely to this and seems to track well with my fitness.

I think it really needs the Garmin recorded data, same as the auto FTP detection etc. It doesn’t get the same quality of information from the TR ride sent across.

FWIW my Garmin 530 bike computer gives consistent estimates with which is based on recent 5 minute power, and WKO5 when I have enough hard efforts (short, medium, long). All three are using different ways to estimate, and all three give the same number +/- 1.

Hmm my is about 5-7 ml/kg/min higher than my garmin estimate. As you said, they use MAP to estimate(w/o HR) Five-Minute Power-Based Test to Predict Maximal Oxygen Consumption in Road Cycling in: International Journal of Sports Physiology and Performance Volume 17 Issue 1 (2021), I don’t have access even with academic license, anyone got it?
and Garmin uses the firstbeat algorithm with HR

Garmin uses Firstbeat

WKO requires an power curve with max efforts short/medium/long.

Intervals uses that paper where VO2max = 16.6 + (8.87 × 5-min relative power output). Example: for 80kg rider and 5-min power = 320W:
VO2max = 16.6 + (8.87 x 320/80)
VO2max = 52.1

Yea I’m trying to read the discussion and caveats, the methods synopsis doesn’t provide much info about the variability of the cyclists and the age component

This also reports about 2ml/kg/min higher than hunter allen’s estimation Hunter Allen Power Blog: How to Calculate Your Own VO2Max

The ACSM formula on Allen’s blog underestimates mine by a LOT. To be within 1 ml/kg/min of the other estimates, I’d have to either gain 60W on my FTP, or keep ftp and lose 15kg.

I have done real testing across the years, and my Garmin has always been the exact same or within 2.


The formula intervals uses based on the FTP I am training to for the TT position calculates 10parts lower than GC. I feel its more realistic though :neutral_face: