The holiday period has consistently been a significant time of growth for VR and holiday 2021 was no different. Largely driven by an influx of new Quest 2 users, VR headsets on Steam handily reached a new record high with an estimated 3.4 million monthly-connected headsets.

Monthly-connected VR Headsets on Steam

Each month Valve collects info from Steam users to determine some baseline statistics about what kind of hardware and software is used by the platform’s population, and to see how things are changing over time, including the use of VR headsets.

The data shared in the survey represents the number of headsets connected to Steam over a given month, so we call the resulting figure ‘monthly-connected headsets’ for clarity; it’s the closest official figure there is to ‘monthly active VR users’ on Steam, with the caveat that it only tells us how many VR headsets were connected, not how many were actually used.

While Valve’s data is a useful way see which headsets are most popular on Steam, the trend of monthly-connected headsets is obfuscated because the data is given exclusively as percentages relative to Steam’s population—which itself is an unstated and constantly fluctuating figure.

To demystify the data Road to VR maintains a model, based on the historical survey data along with official data points directly from Valve and Steam, which aims to correct for Steam’s changing population and estimate the actual count—not the percent—of headsets being used on Steam.

According to the latest data from Valve, VR headsets on Steam reached 2.14% of the Steam audience in January 2022. While this proportion falls short of the 2.31% record reached in May 2021, the estimated count of actual VR headsets has reached a significant new milestone of 3.4 million in January 2022 (compared to 2.98 million in May 2021).

The count of VR headsets on Steam had flirted with the 3 million milestone on several occasions but hadn’t quite surpassed it last month when it saw a significant jump from 1.93% (2.95 million) in December 2021 to 2.14% (3.4 million) in January 2022.

Year-over-year, the count of monthly-connected VR headsets on Steam is up 29.46%.

It’s impressive to see the number of VR headsets on Steam reaching this new milestone, though perhaps even more interesting will be to watch and see how many players stick around for the long haul. Will this be a temporary spike or a lasting impact?

Share of VR Headsets on Steam

Quest 2 was far and away the most significant driver in the growth of monthly-connected VR headsets on Steam in January. Clearly spurred by the 2021 holiday season, Quest 2 reached a record high of 46.02% (+6.40%), its largest single month of growth ever.

Given Quest 2’s significant growth, most other headsets gave up some share of the pie. The Oculus Rift S and Valve’s Index headset saw the most significant losses in share, with falling to 13.10% (−1.67%) and 14.36% (−1.43%), respectively.

Less popular headsets like the original HTC Vive at 7.31% (−0.95%) and Windows VR headsets at 4.99% (−0.70%) lost less share. This suggests that the 2021 holiday season saw many Rift S and Valve Index owners trading up to a Quest 2 while other headsets were less likely to jump on board with the newer headset.

However, despite Rift S, Index, HTC Vive, and Windows VR losing share, all of these headsets actually gained in count. That’s because the size of the pie itself grew enough to more than cover the losses in share.

From a vendor standpoint, Meta’s steady gains accelerated in January thanks to Quest 2 (and despite losses in share from Quest 1 and Rift S), with the company’s headsets now making up 67.25% (+3.67%) of monthly-connected VR headsets on Steam.

This article may contain affiliate links. If you click an affiliate link and buy a product we may receive a small commission which helps support the publication. See here for more information.


  • Chad Nikolaus

    And that’s for PCVR, with a ton of people playing stand alone Quest. I think it’s time for big time devs to get into VR.

    • ViRGiN

      Lmao nobody is interested developing for crowd that waits for -80% steam sales and continues to play the very same 5 games for years.

      • kool

        You do know pc is the biggest gaming market…

        • ViRGiN

          You do know that pc market isn’t exactly pcvr market?

          And I’m fairly sure it’s the mobile phone gaming market that is the biggest.

          • kool

            You described the pc market and moblie gaming isn’t exactly gaming.

          • Andrew Jakobs

            Uhm, playing games on mobile IS gaming. Even a game of patience is a game.

          • ViRGiN

            If mobile gaming isn’t gaming, then PCVR is commodore grade indie 8bit demos.
            You have games like cany crush on one spectrum, and games like alien isolation, call of duty, apex legends and even more coming.

          • Cless

            Very true, the biggest by quite a lot.

        • Christian Schildwaechter

          2021 gaming market:
          – USD 93.2bn Mobile
          – USD 50.4bn Console
          – USD 36.7bn PC

          Mobile gaming is the fastest growing and most profitable sector of gaming, and continues to increase its market share, while the other two have been falling. You can of course define your own market, where mobile isn’t gaming, or Quest 2 isn’t VR, because it only has a mobile SoC, or where the Index isn’t a consumer product, because only enthusiasts can afford it.

          But that is all bullshit, market share is defined by the money flowing, the three areas mentioned include even tinier niches (tablets = mobile, browser games = PC). Overall PC hasn’t dominated for quite some time.

          • NL_VR

            We all know why mobile gaming has big market value.
            I dont think anyone here want to go down that route.
            VR, Pc and console gaming become like mobile gaming = im out

          • Christian Schildwaechter

            Well, Fortnite brings in billions every year with a mobile gaming business model, and Epic is pushing to make Fortnite its entry drug for the metaverse, with concerts and events and now user created virtual spaces called “Party Worlds”. It pays for all the Epic Store giveaways with lots of money left, and I see no way that other big players will not try to create similar money making machines. Meta already tried to bring ads to VR games and most likely only temporarily backed down, because the possible earnings weren’t worth the user outrage at the current market size. But XR/VR multiplayer/metaverse going for services, micro-transactions and ads is a pretty safe bet.

          • NL_VR

            The mobile business model is not fortnite.

            Its more like play for 30min, get to a “paywall”, pay or wait 24 hours.
            Or
            Sorry you died buy extra life here.
            Or
            Play to earn gold and buy new cool stuff… wait a minute, to buy the “cool and usefull” stuff you must earn this other cufrency (diamonds). Earn them by buying this loot chest. Hey hey hey, buy this for $59 It got “most value”

            In other words, mobile games business model sucks.

          • Cragheart

            Mobile gaming is huge, because most people are stupid and pay for those microtransactions and bad gameplay.

          • kool

            By that logic gambling is the biggest market in gaming. Mobile isnt exactly gaming, because its primary business it’s not games it’s the mtx. I know money is the primary goal of any business. In the gaming market you make a game sell it make money. The mobile market is gaas, which isnt gaming it’s a service that provides mtx. It’s like saying novels and magazines are the same market because they both have words.

          • Christian Schildwaechter

            I agree that a lot of the money made in mobile games comes from predatory monetization, and that a lot of this should be illegal. But so far the only thing that was actually prohibited were the EA loot boxes in the PC/console versions of FIFA in Belgium, which qualifies as gambling because that you do not know what you get for your money. In contrast buying 1000 in-game rubies for real money to build a new defense tower right now doesn’t, because you know exactly what you’ll get.

            Before King was bought by Activision, I talked to their data analysts about their data driven level design for Candy Crush Sag. It was both impressive and scary, and I’d consider King one of the more responsible players. But that is not all of mobile gaming. For one there are a large number of cheap indie titles that have no additional monetization at all. And even the most popular free to play titles CAN be played for free, as you’ll save about USD 0.25-1.50 for each hour of grinding. A miserable hourly wage, nonetheless this opened the gaming market to large numbers of new players who are cheap and bad at math.

            The other differentiation usually mentioned is graphics complexity, but that isn’t true anymore either. With modern smartphones featuring very powerful GPUs, they are now running visually impressive (PC port) titles like ARK: Survival Evolved, PUBG or Real Racing. At the same time graphically primitive Undertale is still one of the the best PC RPGs with 50% more ratings on Steam than Skyrim. Undertale was ported to the PSP from 2011, and any modern phone could run it.

            The difference between mobile, console and PC has shrunk, and with game streaming they will merge even more. The business models also tend to merge. Most of the F2P, whale fishing still happens on Android, much less on iOS, but one of the reasons Microsoft just bought Activision Blizzard would be that they are now focusing more on games as a service, fitting nicely into the Xbox game pass strategy.

            That is why I balk whenever someone declares mobile gaming as non gaming. The boundaries have become fuzzy a long time ago, and in the end gaming does not imply a specific technology, monetization scheme, company sizes or input methods. Mobile/console/PC are not equal with different scopes, required equipment, context of gaming, session lengths and different target groups. But whether you spend 50h in Candy Crush on a touchscreen, Fortnite with a gamepad or Fallout New Vegas with keyboard and mouse, in the end all you did was play a game. And hopefully enjoyed that more than the money you could have made by working for 50h instead.

          • kool

            I don’t think you get what I’m talking about. We weren’t talking about playing games we were talking about markets. If I start talking about the real estate market and you start talking rentals, we’re not talking about the same market even tho both of those markets are about homes. The pricing structure, utility and ownership are totally different.

          • Christian Schildwaechter

            You:

            We weren’t talking about playing games we were talking about markets.

            Also you, responding to ViRGiN who mentioned that the mobile gaming market is larger, after you had declared pc is the biggest gaming market:

            You described the pc market and moblie gaming isn’t exactly gaming.

            Not sure how I could have interpreted this in any other way then mobile games not qualifying as games when compared to PC games, as you explicitly rejected that the mobile game market counts when comparing size with your argument.

          • kool

            I’m not sure how you can interpret any other way…the word market is in both comments you quoted.

    • Ad

      Eh, this or the switch? Why start with an entirely new medium?

  • Christian Schildwaechter

    I really wish you would drop that exponential trend line, as it really gives the wrong impression. For one you still treat the development as continuous, which it clearly isn’t. We see huge jumps, drops and seasonal data. HL:A single handedly doubled the numbers in two months, so including this one time event distorts the monthly growth rate way up. Unless you have some info about HL:A Episode 2 you haven’t told us about.

    And now we also have the yearly Christmas cycle, where the numbers jump up in January thanks to all the people who just got a Quest as a present, but then don’t continue to rise at the same rate, but slowly for a few months, when a significant number of the new users looses interest and stops using VR. So ending on January again distorts the trend. If you want to see realistic growth numbers, create a trend line before HL:A, a trend line after HL:A, or even better, trend lines from January to December and compare the years.

    The following graph contains the RoadToVR data and trend line from January 2019 (left lower yellow background) overlaying the current one from January 2022, plus the current (R2=0.928) trend line continued and eye balled linear trends for before (green) and after HL:A (orange). You may notice the 2019 trend line shooting up even earlier than the 2022 one, and none of them matching reality, because they both falsely imply an exponential growth trend.
    https://uploads.disquscdn.com/images/55d799b76e88b6300bcbf1fa47413b36a28ff834ed75f18b60ce3c708577c290.png

    • benz145

      The exponential line is not to imply that the trend is necessarily exponential; it’s to measure how close the trend is or is not to exponential. If it falls under 0.9 I won’t include it in the chart.

      • Christian Schildwaechter

        Again, it only seems to somewhat fit because you continue to ignore the HL:A jump. And you do imply that the trend is exponential by always including a badly fitting exponential trend line while never including much better fitting linear trend lines. You chose this presentation, it is not forced upon you, and you ignore that it has failed to project the actual trend miserably since 2016.

        • benz145

          The trend was very close to exponential, even before Alyx launched.

          https://uploads.disquscdn.com/images/00888d1af15703099c5e4a323084c958ffd511e177455e218080ca361730b502.png

          As we can see from the more recent charts, the exponential fit has been falling away. I have been waiting for it to fall under 0.9 before removing the line so that people can see that what was once a good fit is peeling away.

          • Christian Schildwaechter

            The problem is you can approximate any linear trend with an exponential graph in the lower value range, because growth hasn’t really kicked in yet, just like you can approximate a straight line by a section of circle just by choosing a sufficiently large circle. You could approach the graph above by a properly selected sine function, that doesn’t mean that this function would be an appropriate trend line.

            You actually have to check if your 2019 trend line predicts the trend, which it doesn’t. Instead you basically throw away your trend line with every new iteration of the graph, and create a new approximation with the new data, with the consequence that every time your graph is getting flatter. You can continue to do this even if the actual trend is linear for quite some time. The only reason this hasn’t fallen flat on its face yet is the huge HL:A jump that “corrected” your trend line. If you look at your own trend line from January 2019 and the data you collected afterwards, you see that the trend line was already about 50% off in October 2019.

            What you actually do is sort of curve fitting, meaning finding the exponential growth function that most closely matches your data. But that is not a trend line if you simply replace it every time the data changes, you actually have to take a look at your previous trend lines to see if this is a fitting projection. And if it isn’t, you have to switch to one that better predicts the trend.

          • benz145

            If it were up to you, at what R^2 value would you include the line and at which value would you remove the line?

          • Arno van Wingerde

            Given the huge jump by HL:A, which is a single event and therefore hard to fit in a trend line and moreover the lacklustre development of the last year, the suggestion of any kind of trend is rather dubious. Given the erratic growth within a year you might get away against a data point per year, say January 1st of every year, but it is clear that the result would not have much “predictive power”.
            BTW: could you please order the data sets per producer: Occulus, Valve, HTC ?

          • Cl

            I know this post is old, but even if you say “the chart only looks that way because of X”, it doesnt change the facts of the chart. You cant say “if X didnt happen it wouldnt look that way”, because X did happen. You can clearly see the number of headsets growing faster with time even without the line. I havent seen an updated chart to see what it looks like now.

    • Cragheart

      The fact is that exponential growth looks a lot like linear growth at the beginning.

  • Andrew Jakobs

    Hmmm, i have a hard time believing valve has a bigger slice of the pie as HTC with all it’s different headsets. Might it also include index controllers which might blur the number?

    • Ad

      HTC has one headset, the Vive. The others were of no use to consumers who simply went with the index.

      • Andrew Jakobs

        uhh, no they haven’t. They’ve got the vive Pro, Vive Pro Eye, Vive Cosmos, Vive Pro 2, Vive Focus 3 and now the Vive flow.
        I personally know more people with a Vive than with an Index, but I do know a lot of people who bought themselves the Index controllers.
        I enjoy my Vive Pro very much, recently bought the wireless adapter, and had bought the index controllers a year ago or so.
        Wouldn’t trade my Pro for the Index. I do hope there will be a new better headset soon with OLED displays and a larger FOV as my Pro. But then again, with the price of GPU’s to drive these newer headsets I think I’ll wait anyway. Still have a RTX2060super which isn’t powerful enough to drive any of the newer headset as it already has problems driving the Pro.

        • Ad

          These devices are irrelevant.

          • Andrew Jakobs

            How so?

    • benz145

      The vendor breakdowns are official from Valve, and they’ve had the #2 spot for a while. If it wasn’t accurate I think we’d hear something about it from HTC.

    • ViRGiN

      Don’t underestimate brand loyality, and Valve has a very real one.

  • Ad

    How are people still using Vives?

    • ViRGiN

      “people” still use 3 year old index that has never received a single update and they still call it high end, so whats your point?

    • Andrew Jakobs

      It’s more than just the original Vive, it’s also Pro (2) and Cosmos.
      And in the end, if people are satisfied with their original Vive because they don’t have better GPU’s to drive the newer headsets, who cares. It’s still a big immersion.

  • mirak

    My Vive Wireless isn’t detected by Steam if Steam VR isn’t running and the headset on.

    • Andrew Jakobs

      But most people use Wireless with Steam VR don’t they? Don’t know if Viveport is used very much, personally I keep forgetting that I also have it.

  • This continues the trend of the last months: PCVR is growing, but grows slower than standalone market

    • Andrew Jakobs

      It will be a hybrid. Pure PCVR isn’t gonna survive.. People will want standalone (as in wireless) with the option to stream from PC.

  • kontis

    Sorry to say this, but this title is just speculation, so you shouldn’t state it as a fact.

    We don’t have MAU numbers of Steam, at least not any recent ones, so this number cannot be accurately calculated.

    • benz145

      It’s based on our model which we believe is accurate, especially with the overall trend. Even if 3.4 million wasn’t accurate as an absolute count, we can still see that there has been a large increase compared to the month prior.

  • Arno van Wingerde

    This suggests that the 2021 holiday season saw many Rift S and Valve Index owners trading up to a Quest 2 while other headsets were less likely to jump on board with the newer headset.

    However, despite Rift S, Index, HTC Vive, and Windows VR losing share, all of these headsets actually gained in count. That’s because the size of the pie itself grew enough to more than cover the losses in share.

    Hm… that was a bit confusing to me as the first part might suggest that trading up mechanism, which obviously does not occur because their counts still go up: If you would first mention the actual counts of the other VR headsets goes up and then go on to say they still loose share because of the massive growth in Quest 2 headsets, the conclusion gets across easier. Yes, it is less interesting that way, but easier to follow.

  • Cragheart

    Almost 6 years after Oculus Rift CV1 and only about 3 million. That’s not much honestly. Mobile VR is sh*t btw.

  • Cragheart

    Trust me, we don’t want Facebook to monopolize VR or AR.

    • ViRGiN

      LMAO what a non-facebook fanboy.
      Instead of asking for companies to step up, you’re hoping for facebook to stop lol