Any cyclist worth their chamois knows that body weight is a huge performance factor for riders. The more mass you have to move, the more power you have to put out to move it!
And this is true no matter what road you may ride. Sure, weight matters the most on climbs. But it’s also a factor on flats and descents.
We decided to do some testing to see just how much each kilogram of weight slows (or speeds) a rider on Zwift. Our tests included flat roads, climbs, and descents. And they answer that eternal question: should you work to lose weight, or invest your money (Drops) in a faster virtual ride?
We found the results quite interesting, and we think you will, too.
A few notes about our tests:
- All tests used a 183cm (6′) tall rider riding at 300 watts on the Zwift Aero frame with Zwift 32mm carbon wheels.
- We ran each test at two different rider weights: 75kg and 82kg. We went with a 7kg difference because Zwift HQ says this is the weight difference between Zwift’s lightest and heaviest bikes. So the speed differences you see below could be realized purely through equipment changes in game!
- Riding at higher or lower wattage or being significantly heavier or lighter than our test riders would impact time gaps somewhat, but we choose to keep the tests simple so the results would be easy to understand.

Effect of Rider Weight on Climbs
Climbing tests were completed up Alpe du Zwift, an 8.5% average gradient which is 12.2km long.
Test Results
- 82kg @ 300 watts: 54 minutes, 49 seconds (13.4kph)
- 75kg @ 300 watts: 49 minutes, 31 seconds (14.8kph) – 318 seconds faster
Climb Test Observations
Not surprisingly, up the steep Alpe climb is where we can really see how much weight affects rider speed. The time gap here is over 6x what we see on the flats! This is why lightweight riders win mountain races outdoors while the heavier sprinters struggle to finish within the time cutoff.
Conclusion: reducing your body weight by 1kg will save you ~45 seconds up Alpe du Zwift.

Effect of Rider Weight on Flat Roads
Flat tests were completed over two laps of the Tempus Fugit route.
Test Results
- 82kg @ 300 watts: 52 minutes, 17 seconds (39.7kph)
- 75kg @ 300 watts: 51 minutes, 26 seconds (40.4kph) – 51 seconds faster
Flat Test Observations
While some Zwifters like to say “it’s all about pure watts on flat roads,” it’s not that simple. Weight does affect your speed on the flats in Zwift, since it changes your computed CdA. Or to put it another way: two riders of the same height but different weights will have different CdA values in Zwift, meaning the heavier rider will need to push more watts to keep up with the lighter rider, who is encountering less (virtual) air resistance.
Conclusion: reducing your body weight by 1kg will save you ~9 seconds over an hour-long flat race effort on Zwift.

Effect of Rider Weight on Descents
Descent tests were completed down Alpe du Zwift, an -8.5% average gradient which is 12.2km long.
Test Results
- 75kg @ 300 watts: 10 minutes, 7 seconds (72.5kph)
- 82kg @ 300 watts : 10 minutes flat (73.3kph) – 7 seconds faster
Descent Test Observations
Riders complain about getting dropped on descents in Zwift, but that isn’t because the algorithm is computing speed incorrectly. The real reason riders get dropped on Zwift descents is that the pack keeps pedaling on the downhills. If you aren’t putting out power on a downhill in a race, you’ll get dropped–no matter what you weigh. Unless you supertuck smartly.
Conclusion: reducing your body weight by 1kg will cost you 1 second on a 10-minute Alpe du Zwift descent.
Buying Upgrades, Riding Up Grades
The great Eddy Merckx famously said, “Don’t buy upgrades, ride up grades.” He was making the point that, if you want to be a better rider, training is the way forward. Buying more expensive equipment will get you marginal gains, but increasing your power and/or reducing your weight will get you greater gains, without impacting your bank account!
Our test results above coupled with hundreds of additional test laps show an notable “2kg line” in Zwift performance on flat roads:
- The difference between the fastest frame + wheel combo and a basic racing setup (Zwift Aero frame with Zipp 808 wheels) is 18 seconds over an hour ride. This is roughly the same time improvement you would see by dropping 2kg of body weight.
If you’re carrying a few extra pounds and find yourself concerned about your Zwift race bike setup being a bit slow–see if you can shed some weight. Dropping just 2kg (4.4 pounds) will have the same effect as upgrading your basic Zwift ride to a top of the line racer, assuming you’re able to maintain the same power.
There is also a “1kg line” for Zwift performance on steep climbs like the Alpe:
- The difference between the fastest frame/wheel setup on climbs and a basic racing setup (Zwift Aero frame with Zipp 808 wheels) is 52 seconds over an hour climb. This is roughly the same time improvement you would see from dropping 1kg of body weight.
If you still haven’t unlocked the Lightweight Meilenstein wheels atop the Alpe, there’s a way to make up for it: losing just 1kg will save you more time than those wheels ever could. Time to start counting calories, perhaps?
An Important Note
If you’re overweight, losing those unhealthy pounds can improve your cycling performance both on Zwift and outdoors. But any attempts at weight loss should be handled carefully, and ideally under the watchful care of a qualified professional.
In a competitive, weight-based, results-driven environment like bike racing it’s easy to go too far and actually damage your body with extreme dieting or disordered eating.
Read “RED-S: One Athlete’s Story” for more on this topic.
Your Thoughts
Inspired to shed a few pounds? Surprised by the results above? Share any questions or comments below!
Note: This is a simplified version of the original post. The first version was revised after it generated a lot of confusion that distracted from the goal of learning “how much faster you’ll be if you lose 1kg of weight.”
As always, really good work Eric! Thanks.
At 183cm/118kg (6’/260lb) I am the definition of a Clydesdale. Since getting on the bike again and Zwifting, my weight is coming down. As the weight comes off and I ride more and harder, my watts are increasing, and importantly for how distance and speed is calculated in Zwift my w/kg is increasing. Merckx was partially correct – but he uses the wrong conjunction. It isnt an either/or proposition, you want both. Greg LeMond says it best – “It never gets easier, you just go faster.”
If you can drop weight AND increase power, you’re doing it right, Craig. Keep up the good work, and enjoy watching that w/kg number bump up!
It always pleases me to hear of other Clyde cyclists out there; we have unique struggles in the cycling world. I assume from this statement you are not weight doping, and neither am I. I am 185cm/124.7kg (6’1″/275lb) and have not weight doped. Yeah, the “easy” D group rides are a workout for me, but I can hold on. My weight is also slowing dropping and my watts are increasing. Best of luck out there on your journey!
That being said how many people put their true weight into the metrics???
Er, me! I can only hope to achieve moral victories anyway.
I do because I want my data to be able to show my true improvement as I lose weight and get fitter.
Een minder gewicht inzetten is jezelf bedotten 😀
I think the vast majority of folks play fair. I know I do! But sure, there are some weight dopers.
in races I have the feeling that a lot of the riders are cheating, I know for myself that I can ride the speeds outside as I achieve in zwift. Hope zwift will take measures in the future for more faire racing and riding in zwift.
I link my Garmin. Account to Zwift and weight is transferred automatically. I gained four kg last year and my ADZ time slowed by 6 mins (48- to 54) despite putting a lot of work in. I wish I didn’t like beer and carbs so much!
Most honest people do, the only one you cheat in games or like such by cheating is yourself. 🙂 – So is it worth it to use inputs that is not right and get no real satifaction or gain out of it ?
I do lol or what’s the point?
??? Do you? I prefer to think that big majority is honest. Looks like you don’t have faith in humanity!
I’ve just discovered my zwift weight syncs automatically with my fitbit scales, so this is quite good motivation.
Yes I do
Yeah I’m strict with the weight I put in. 180cm and Currently float between 81kg and 83kg over a week. So I put in 82kg. Even if I weigh in at 81 before a ride, I don’t change it until I weigh an average of 81kg over 3 weigh ins through the week…. Zwift is like golf in my opinion. If you cheat, you’re only cheating yourself.
Good point you make about the averaging Dave.
between the start and the end of a hard workout, it’s normal to lose nearly a kilogram in weight. This is not only in sweat but also in what we exhale.
Not to mention the obvious that eating a main meal or going to the toilet, also affects what we see on the scales.
I know it’s an old subject but I just found it together with your comment. I am, at this moment in time based on Zwiftpower lists of my events, probably the heaviest zwifter out there. I am gradually losing weight but I made a resolution to update my record in 5kg decreases. I am close to update it with 5kg less any day now. Whats the point of cheating myself? I am happy if I can finish an event a few places from the bottom. That’s my reward.
Thanks to a DC Rainmaker article, I have an inexpensive scale that links to my phone via Bluetooth which then updates my weight to Zwift. Still a clyde but an accurate clyde.
Great analysis!! And spot on follow-up comparisons utilizing the data to show that the indisputable best upgrade is almost always a loss of a few pounds! As a (current) 82kg rider, who can hang with the very front A/B group on an entire flat race, but gets dropped like a sack of cement out of a four-story window the second we hit any proper climb, this is a sad but true reality. :chuckle:
I’m right there with ya, @Bobby Mac.
Im 81 kg but I have the opposite. I can do 3,0w/kg avg for an hour flat, but on a climb I can do 3.5 to 3,8. Standing I do 4,0 w/kg. So my best results are always on rides/races with climbs
Feel your pain!!!
I read this article with great interest a couple of weeks ago when you included a comparison to the (simulated) real world. Why did you take the comparison out … which I thought was the most interesting piece? Too many death threats from lighter riders? 🙂
A couple reasons, actually. One is that the real world calculators all return different numbers–some that are more in line with Zwift’s speeds and some that are not. So there’s confusion there, and I didn’t want to paint the picture that Zwift gave an unfair advantage to any particular set of riders.
But the bigger reason was that I thought it distracted from my goal of the post, which was simply to show how much faster you’d be on Zwift if you lost 1kg.
Thanks. Really appreciate your thoughts. Still … I think your original idea was very intriguing
Glad I saw this comment, I’ve just spend around 60 minutes trying to find the article where you compare Zwift rider weight speeds with the real world!
I’m not a big believer in the Mandela Effect but was seriously starting to doubt my own sanity when I could only find this article, which looked exactly the same, yet for some reason missed the key lines I was looking for.
Unfortunately I’m still a 90kg rider on Zwift, though on the plus side, appears I haven’t quite gone completely insane just yet!
… yet. Ha!
I feel for you Duncan, I’m the same… I can’t give up when I know there is something somewhere in the cloud that I have seen or even wrote myself before, and now I can’t find it. I’ve spent far more time looking for it than it would take to simply write it again.
Would be a lot more interesting/pertinent if the test was based on w/kg since Zwift categories are based on w/kg. Comparing 300w @ 82kg = 3.65w/kg and 300w @ 75kg= 4w/kg would expect the higher w/kg to be faster. However, what is the time difference on flat/climb/downhill between a 82kg @ 3.25w vs 75kg @ 3.25w?
I’m working on a post along those lines, actually. We’ll see how long it takes for me to finish!
Great! This is what I was looking for while reading the article.
Any update on this? I drive at 2.5w/kg, but weighing in at 100kg I really struggle uphill compared to lighter riders doing the same w/kg.
You should go faster than lighter riders doing the same w/kg (see Zwift Insider article called speed tests at equal w/kg or something like that). Obviously depending on your FTP and their FTP if it is a long climb the other riders might be able to hold power for longer, but you should go the same speed or faster for equal w/kg on climbs as well as flats
Hi there. I was wondering if finally you post that article. Becuase, since riding in Zwift, I wonder if the w/kg eliminated the weight contribution, or the light one had an advantage, probably becuase it would have less aero resistance.
The phrase in the excellent article that needs re-read by all who want to gain speed by losing body weight is: “assuming you’re able to maintain the same power”. Can those who are not returning from layoff or injury lose 4.5kg or so and maintain the same power? I would like to hear from those who have accomplished that.
That’s always the hard part, isn’t it?
Currently working on my weight. I personally won’t weight dope as I want MY results on Strava/Garmin Connect to be capable of showing MY personal improvement. I don’t race but group rides are tough to keep up even the slow ones due to my weight. It gets suggested to weight dope so can stay with group. I want my results and PR’s to be mine and not because I digitally lost weight. I just recently lost 1kg+ as I am working on my diet and weight. Thanks for the awesome info as always.
Keep up the good work, @Shawn Schmitz!
good test but if you tested more than 2 weights it might be interesting to see if the relationship is linear or curved, is it really as simple as saving X seconds per kg or is it a law of diminishing returns or does it actually increase the effect the more you lose?
It’s a pretty linear thing, actually. I just kept the data simple, but all the way from 50kg-100kg it’s a pretty linear thing.
So what you’re saying is that I’m fat.
You said it, not me! 😯
Good stuff Eric! I enjoy reading your posts. 🤙🏻
Thanks!
Thanks for this Eric, interesting indeed. I realise it’s not something that can (easily) be altered IRL but I’d love to know the difference an individual’s height made to speed/time.
Here you go, @guest_30447: https://zwiftinsider.com/how-height-affects-speed-in-zwift/
I am not a physicist but I did stay at a Holiday Inn Express – it might be interesting to look at descending as well.
I did. See “effect of rider weight on descents” above. 😉
If people weren’t “inspired” to weight dope already, this will inspire them 🙂
Having been at 68kg +/- 1kg for around 40 years I have just managed to hit 67kg with a target of 66kg for a real world hilly 25TT.
Good to know that I will have some gains in the virtual world from my weight loss efforts !!
I’d say cheers but one of the requirements is no alcohol – empty calories 🙂
Hello everyone at Zwift
I really enjoy your website and am an avid participant .
I would like to request if you could set up races or training for 65 years and over Riders .
Sometime it is difficult to pair up with the younger generation !
Thank you
Mike V.
Very interesting. But you don’t mention that heavier riders usually produce more watts. Naturally heavier on the pedals and bigger muscular mass. As a lighter one, I always prefer hill where it’s a matter of w/kg. On flat I struggle, having to produce much higher w/kg than most other rider…
Question is of course, do bigger riders produce more power simply because they’re bigger or are riders bigger because they produce more power?
Personally, as someone who’s spent a fair amount of my youth in the gym doing strength work and eating correctly, I find it interesting in Zwift that smaller riders seem to think extra power is just something that’s natural with weight rather than earned.
There’s always an option to work hard to add weight as well as to lose it.
I had lost about 5 kg over the past several months but was reluctant to change my avatars weight thinking I was faster in flats so you can bet I’m changing my in game weight to reflect my true weight based on this info.
Hi Eric I have noticed that my avatar is a bit porky love handles etc. If i was to lose some weight will my cycling screen buddy look a bit neater as well.
There are three avatar body types, based on BMI thresholds. See https://zwiftinsider.com/avatar-sizing/ for details.
Eric, I have been re-reading this article with added interest. This is in order to understand why I have been struggling to compete against heavier riders during my recent flat TT events. I have experienced heavier riders maintaining a higher speed on the flat due to being able to produce more watts and still not exceeding the overall category limit (Cat C – 3.2w/kg). I wonder if you did the tests using the category limit (3.2w/kg) as the static value and not for a set power value this would show more of a difference. So for a 75kg rider @… Read more »
Yes: if you’re a lighter weight rider, you’ll perform better vs the competition when there are climbs involved, assuming you and they are both near the category w/kg limit.
Thanks, it’s so easy to get your understand confused when trying to come to terms with all the variables as well. Really appreciate the articles and information you publish, clear and concise. Looks like it’s off to Bologna for me.
Cheers, Andy.
Eric … can you supplement this article to include a comparison test at a static wkg instead of absolute 300w output. On a unscientific test I did, the heavier rider outperform the lighter rider on flats and climbs when riding at the same wkg.
Already done, Remy! See https://zwiftinsider.com/wkg-tests/
Dear Eric, thank you. I just started in Zwift and I have a (stupid) question. Should I enter my ‘naked weight’ or include the weight of my bycicle?
Naked weight. Your bike weight will be added to that (and it varies depending on which bike/wheelset you use!)
I love zwift, but weight pens on the alpe is unfair as it is not true climbing…IRL climbing is much much harder as you really are carrying your weight, O n zwift you are not doing this..it does not matter if you are 60kgs or 100kgs! So if I took 5kgs off my profile weight, my time up the Alpe would be nearly 4 mins quicker?
Eric, really appreciate all the info you provide. However, having read all about how uphill speed is calculated I still can’t understand how a rider putting out 2.1 w/kg on the ‘bonus climb’ was riding away from me convincingly as I put of 3.1… no kit seems to have that much impact.
Do you have any insights?
Thanks!
Each route has a characteristic joule/kg. Is this available anywhere?
I am surprised the speed improvement on L’Alpe was apparently 10.7%, while the change in (inverse) weight was only 9.3% And that just includes rider mass. If I add in an assumed 6.8 kg for the bike, 2 kg for clothes and such, the weight advantage is down to 8.4%. And that doesn’t include wind resistance, which would further reduce the speed advantage. I can’t think of any physics in the standard power-speed model which would cause speed to increase “faster” than weight changes. Here’s how I’ve personally treated height, BMI, and wind resistance: If I assume CdA is proportional… Read more »
How can I tackle the descents as quickly as possible?
Thanks and good week.
Just to confirm, the w/kg that zwift displays and uses, is based on the weight of the rider plus the weight of the bike he is riding, correct?
No. w/kg never includes the weight of the bike, even when you talk about it outdoors. Just rider weight.
Something that’s always confused me…when entering weight, should it include being fully kitted up in shoes, or in your birthday suit…ahem..? Does Zwift account for the weight of shoes for example when calculating CdA? Just wondering what the standard weighing process is?
All the Zwift clothing has weight included. So I weight myself in my birthday suit, or just a kit…
Normally my weight is automatically uploaded to Zwift every morning from my Withings scale. The other day I weighed on a regular scale and went to manually input my weight and noticed that you can only use whole numbers. Does that mean that Zwift is always rounding my weight to the nearest whole Kg even though Withings provides a more precise number? ZwiftPower seems to use the more precise number. No big deal, but it just got me thinking :).
Hey Eric, this is a great article. I’m unclear if you can derive from these tests what two riders of different weights, pushing the same w/kg will end up doing and who would be fastest?
Your test of the lighter rider pushing the same, 300 watts, as the heavier rider, resulted in the lighter rider being quicker, with higher w/kg. Would be interesting to see the same test matching w/kg (as opposed to pure watts), maybe you’ve done this already?
Yes, see https://zwiftinsider.com/wkg-tests/
Hi Eric — any idea if weight (rider + bike + wheel) affects acceleration in zwift? I feel like I accelerate through packs more quickly with hard efforts on a light bike than a more aero bike. But it might just be in my head.
I would think it does, although it’s hard to test because we just don’t have the granularity in the data that we need. Now, whether you’ll feel much of a difference between two different bikes… I think that’s probably more in your head than anything else. But that doesn’t make it a bad thing. 🙂
I think there is a problem with the fact that weight isn’t factored around BMI or body fat percentage. Coming from bodybuilding I weigh 200 but can leg press over 1000 and can attain around 16 to 17 miles per hour sustained for 20 miles real world. Last week did a 1 mile climb at 8 percent grade and never went below 4mph. First night on zwift and my leg are moving furiously in big gear front and middle gear back on what seemed like a small hill. Zwift said 2mph….
Hi Eric, just wanted to know if zwift takes bike weight into the equation
I estimated from another empirical model that the time savings are between 200-250seconds on an average 8.5% slope so there’s nearly a minute extra in your analysis. Are you setting this up on an emulator? Good article anyway…
Using an ANT+ simulator to send our bot up the Alpe, yes. Did you do your tests without any drafting? We do ours solo, so the rider doesn’t interact with any others that would influence his speed…
Hi Eric, i am ex track cyclist who has put on a (quite) few pounds. I have found your analysis really good and interesting. My wife has had enough of me whinging about me not being able to stay with any of the Zwift D races! I have lost kg since getting back on the bike and feeling much stronger. It is getting really frustrating not being able to keep up. Last race I started at 350w everyone rode away from me. At 350w I still do not record a 3 kg/w ratio. Wife says just change weight to 70kg,… Read more »
Eric,
I am new in Zwift and my apologies for my question in advance. So based on this post, yes, it is true that Zwift take your weight in consideration when climbing to simulate your real climb
Absolutely true!
Hi Eric My partner and I zwift ‘together’. He’s 6’1″ weighing 213lb and I’m 5’4″ weighing 129lb. When he cycles on zwift with his true stats loaded he trails me, but when he fibs and puts his stats the same as mine, he disappears into the distance ahead of me. He thinks that by putting our stats the same, he puts us on an even playing field…. is this true? or does his actual body mass on a static trainer give him an advantage over me because of his weight down through the pedals? Just to add… out on the… Read more »
So to be exact, do I need to type my weight in the morning after peeing into zwift or is it the weight that I have on the bike (how would you know that??)
It’s just your own body weight. Bike weight, etc is added to that by Zwift’s physics engine.
Fantastic article. All my mates buying carbon race bikes for 8000 pounds should look into getting rid of their monstrous beer guts first! But buying a bike lots easier than losing weight. Lol!
Hello everybody! My experience with this is somehow different. I’m a quite heavy but strong rider (108kg and ~390 watts ftp). Today I raced the ocean lava cliff side route (Cat. A) and was easily able to hold on to the 2nd group in the steep climb (first part of epic KOM reverse). When I take a look at the race results I always have the lowest avg w/kg of all the riders around me… Any thoughts on this?
Hi Eric, I’m new to Zwift and have been concentrating on hilly routes, at 89 kilos I really suffer going uphill as expected but I think I maybe doing it all wrong. I rode Epic Kim today with my cadence 60 to 70 rpm, would my overall time differ if I spin the gear a bit more?
What about for us LIGHT riders that are getting smoked on the flats does it make sense for us to go heavier on frames?
If you’re getting smoked on the flats, it makes sense to go with the most aero frame possible. And become really good at drafting…
I’m 6’6″ and 90kg. This means I’m getting heavily penalised for both height and weight.
I don’t think the Zwift calculation works well for people like me, as I find I’m being dropped easily by riders in Zwift that I can out-climb and out-race in the real world.
Any advice, except losing weight 😂
Thanks for creating this article Eric. So now I know why I beat all my mates up hills in Zwift and conversely get dropped downhill without peddling hard
I’m genuinely 65 kg and 5’9″ with an FTP of 2.4 w/kg. Out on the road my buddies all pass me going uphill (no stamina, me) but I go downhill quicker (presumably cos I’m more aerodynamic. Completely opposite to my mates. So it’s nice to beat them even if it isn’t real.
Taking an idea from your drafting tests, do you have any thought of running tests at 200 or 225w?
As a long-term D rider, I suspect without data that the benefits of super-aero bikes are proportionally less at lower speeds AND weight penalties are proportionally greater.
Not sure if this was discussed in the comments. But a more interesting perspective is how much of a difference riders of different weight, gain or lose time while riding the same w/kg up a climb or on the flat. Or how much more savings does a lighter rider gain while slipstreaming as compared to a heavier rider.
Thank you for this good information. Btw, do you have this published on a journal, conference or books? Thanks in advance!
Great article Eric, when I started to use Zwift I was around 77kg and after a few months riding it dropped steadily to around 75kg and held there, however I stopped drinking some 8 months ago (just decided to stop, no real reason, just one day said no thanks) and the weight dropped off to a now very steady 69kg with no dieting or extra exercise. it really didn’t bother me not having a beer/wine and was no hardship at all, I am amazed that the alcohol made such a weight difference. I may return to having a cold beer… Read more »
Hi Eric. Great work. I would like to see a bit different test: watt required to uphold 40 km/h (perhaps also 35 and 45) as a function of rider weight on the flat (perhaps also grade 5 and 10). That is, the x-axis would be rider weight and the y-axis would be watt (perhaps an additional interesting plot would be a different y-axis with rider watt/kg). Hope you find it interesting enough to do. Thanks
Where did the 1kg conclusion numbers come from and what justifies them? Is the power, gradient, weight -> speed function linear? If it’s not, which I don’t think it is, you can’t just say well a 7kg difference is 318 seconds so a 1kg difference is 318/7=45s difference. That works only if the function is linear in mass. But acceleration is linear in mass, which means velocity is squared since it integrates acceleration. This means that the difference in speed in the first kg increase in weight will be less than the difference between the 6th and 7th kgs in… Read more »
In an AutoCat race I saw a 35kg rider with a 4.5w/kg FTP, who was placed in Cat D. Was wondering whether he is really equivalent to a 70kg rider with a 2.5w/kg FTP.
Lower in the comments found your article https://zwiftinsider.com/wkg-tests/ which explains a lot. Thanks!
Hi Eric, I wondered if you had thought about redoing the two “heavy” tests (and maybe even the height ones as well) given the changes in rider dynamics over the last couple of yeas?