Video Game Emissions Study: What The Numbers Actually Say

A 2026 video game emissions study estimates that gaming produces 62.51 million tonnes of CO₂-equivalent a year across the categories it counts. Using 2024–2025 data, the model estimates that PC play alone accounts for 9.779 million tonnes. That’s a striking total, but it is not a direct measurement of every console, gaming PC, or player. No one followed every gamer around with a tiny carbon clipboard.

The boundary matters. The study excludes mobile gaming, physical-copy distribution, recycling and disposal, esports events, merchandise, and game streaming. It also relies on proxies and assumptions where data are missing, including assumptions about downloads and playtime. So the headline deserves attention, not blind faith. The useful question is not just how big the number is, but what the model counts to get there and what it leaves on the cutting-room floor. The paper’s methods and exclusions make that distinction hard to ignore.

Key Takeaways

• The 2026 study models 62.51 million tonnes of CO₂-equivalent in annual gaming emissions from 2024–2025 data, but excludes mobile gaming, physical-copy distribution, recycling and disposal, esports, merchandise, and game streaming, so the headline is not gaming’s complete footprint.
• Hardware manufacturing deserves a seat at the table: the study estimates 11.52 million tonnes of CO₂-equivalent for PC manufacturing, more than the 9.779 million tonnes estimated for PC play, according to the paper’s category estimates.
• No delivery format wins every matchup: a 2020 University of Surrey study estimated 0.047 kg CO₂-equivalent per gameplay hour for downloads, 0.055 kg for discs, and 0.149 kg for cloud gaming, while its modeled duration comparisons sometimes favored cloud gaming for shorter play, as the University of Surrey study found.
• Gaming-emissions claims deserve trust only when the boundary, evidence, and assumptions are clear, so check the study’s methods before repeating its headline or treating a modeled estimate as a personal carbon score.

What Does 62.51 Million Tonnes Mean?

The 2026 paper estimates annual video-game-related emissions at 62.51 million tonnes of CO₂-equivalent across the categories it includes. That’s a modeled estimate based on 2024–2025 data, not a comprehensive direct measurement of gaming’s entire footprint or a precise tally for every player. The headline is big, and the fine print is doing important work too. The study is a calculation assembled from available evidence, not a planet-sized emissions meter.

CO₂-equivalent puts different greenhouse gases into a common unit so their warming impact can be compared. It does not mean researchers measured 62.51 million tonnes of carbon dioxide coming directly from games, consoles, or players. The estimate combines emissions across modeled parts of the industry, so the total depends on what the researchers counted and how they turned incomplete information into numbers.

The researchers draw on public company reports, hardware and sales data, plus estimates of game downloads and playtime. Those sources do not provide a complete, consistent record of every relevant activity. Where the data run out, proxies and assumptions step in. That is not automatically a flaw. It is how researchers can build an estimate when direct records do not exist. But the result is only as sturdy as the inputs and choices behind it. The paper’s methods explain how those pieces feed the model.

Don’t mistake the precise-looking “62.51” for precision about every underlying activity. A modeled total can show the scale of a problem while still carrying uncertainty. The decimal places are arithmetic, not a promise that the real-world footprint is known down to the last fraction of a tonne.

So the number answers a narrower question than “What is gaming’s complete climate impact?” It estimates annual emissions associated with the categories the researchers modeled, using the evidence available to them. Any gaming-related activity outside that boundary is outside the total too. That distinction turns a dramatic headline into something more useful: a defined estimate with limits readers can actually see.

Where Does Gaming’s Footprint Come From?

The modeled footprint includes hardware manufacturing, electricity used to play, game development and publishing, cloud gaming, and digital delivery. The paper estimates annual emissions for each category, but those figures come from different types of evidence and assumptions, not one giant meter attached to the games industry.

The paper’s manufacturing estimates, category estimates, and Steam download estimate are all modeled figures. The study estimates that Steam transferred 469,119 petabytes of data in 2024, then associates that volume with 4.38 million tonnes of CO₂-equivalent. A data-transfer estimate is not the same kind of quantity as a manufacturing estimate or an estimate of electricity used during play. The table maps the paper’s categories; it does not present perfectly measured, interchangeable slices. The researchers combine company reporting with hardware data and assumptions about activity where direct records are missing. PC-play emissions, for example, are extrapolated from estimated Steam player-hours. Cloud gaming estimates rely on assumptions about users, playtime, and data-center rendering hardware. These results help show what the model considers significant. They do not mean every category was directly metered in the same way. Manufacturing belongs in the discussion alongside electricity use because a gaming device has a footprint before anyone boots up a game. The paper estimates 11.52 million tonnes of annual emissions for PC manufacturing, more than its 9.779 million-tonne estimate for PC play. Those figures cover different parts of the system, but they make one point hard to dodge: judging gaming only by the power drawn during a session misses emissions from making the hardware that session depends on. The same distinction applies to consoles and monitors. Their manufacturing estimates are annual category totals, not emissions assigned to each player or hour of play. A clean-looking electricity bill does not make the production stage disappear. It measures a different part of gaming’s footprint.

What Did Researchers Leave Out?

The researchers left several gaming activities outside the initial total and used proxies and assumptions to estimate some activities they did include. Exclusions set the boundary of the headline; proxies shape the numbers inside it. According to the study’s stated exclusions, hardware assumptions, and download methodology, the initial total does not include mobile gaming, recycling and disposal, physical-copy distribution, esports events, merchandise, or game streaming. So the figure does not cover the whole gaming ecosystem. This is not a complete inventory with a few tiny side quests missing. For gaming-PC manufacturing, the authors use two stand-ins: a Dell G15 laptop with an adopted manufacturing-emissions estimate of 308 kg CO₂-equivalent, and a Dell Vostro 3030 desktop at 112 kg for manufacturing and transport. The hardware assumptions use specific products as proxies for broader PC categories. A proxy is a practical substitute when data are incomplete, not a measurement of every gaming laptop or desktop. The hardware mix chosen for a model can affect its estimate. The digital-download calculation also depends on set inputs. The authors assume 20 watt-hours of energy per gigabyte transferred and a global grid carbon intensity of 0.445 kg CO₂-equivalent per kilowatt-hour. Multiply data volume by assumed energy use, then apply the grid factor, and you get an emissions estimate. That is a transparent method, not a record of the electricity used for every download in every region. These limits matter in different ways. Excluded activities fall outside the total entirely. Proxy choices and calculation assumptions influence estimates for activities inside it. Treating both as one vague “caveat” blurs a distinction the paper’s accounting needs readers to see. The omissions do not prove the paper is useless, and they do not tell us the real-world total must be higher by some known amount. The study provides no measured emissions add-on for the excluded activities, so assigning one would be another guess dressed up as arithmetic.

Are Digital Games Always Greener?

No. Digital games are not always greener. The answer changes with delivery method, game size, and how long someone plays. “Digital is always greener” is a slogan, not a finding that holds up under every set of assumptions. A 2020 University of Surrey life-cycle study estimated average emissions per gameplay hour at 0.047 kg CO₂-equivalent for downloaded console games, 0.055 kg for disc games, and 0.149 kg for cloud gaming. Those averages put downloads ahead, but the study’s modeled comparisons show why one ranking cannot settle every case. The University of Surrey study found that cloud gaming could have lower estimated emissions than downloads when a game was played for under eight hours, and lower emissions than disc games for play durations up to 24 hours.

The apparent twist is about the comparison, not a maths error: the hourly averages and play-duration thresholds answer different modeled questions. The Surrey results depend on their assumptions, including how emissions are allocated across a game’s delivery and use. Playtime can change which option comes out lower. Game size can flip the result too. A 2014 life-cycle study examined an 8.8-gigabyte PlayStation 3 game using 2010 data and found that physical Blu-ray distribution had lower estimated emissions than downloading for games larger than 1.3 gigabytes. The Journal of Industrial Ecology study is an older, specific comparison, not a universal verdict on discs.

That result does not directly contradict the Surrey study. They use different data, assumptions, and comparisons: one examines a particular PS3 game and its distribution, while the other models console delivery options against gameplay duration. Neither gives a permanent winner for every game, player, and delivery route. The newer global estimate asks a broader question about annual gaming emissions across selected industry categories. It is not a head-to-head life-cycle comparison of physical and digital copies. Different studies can produce different answers because they count different things, not because one has to be wrong.

What Can Older Gaming-Energy Studies Tell Us?

Older studies show how much electricity gaming systems used in the United States, but they do not form a neat historical trend line for the newer global emissions estimate. The 2019 Lawrence Berkeley National Laboratory study measured 26 gaming systems and estimated that gaming used 34 terawatt-hours of electricity and cost $5 billion in annual energy expenditures in the United States. Those are striking numbers, but they describe a U.S. estimate, not worldwide gaming emissions. Lawrence Berkeley National Laboratory’s 2019 study

An earlier LBNL field study collected data from 880 U.S. households in 2012, including readings from 113 game consoles. Researchers estimated that U.S. consoles used 7.1 terawatt-hours of electricity that year. The study is useful evidence about console use in American homes. It is not a global count of every gaming device, much less a measurement of emissions from the entire industry. The LBNL field study

The differences are not minor bookkeeping. One study estimates U.S. gaming electricity from measured systems; the other estimates U.S. console electricity from household field data. Both are tied to particular years and a particular geography. Electricity use is also not the same unit as CO₂-equivalent emissions. Converting power consumption into emissions depends on the electricity supply. The newer global model asks a much broader question. Using 2024–2025 data and modeled inputs, it estimates annual emissions across categories including device manufacturing, PC play, game development and publishing, downloads, and cloud gaming. The 2026 study is therefore not a sequel to the U.S. console studies, and its broader scope does not make their numbers outdated. Putting 7.1 or 34 terawatt-hours beside a global total in tonnes of CO₂-equivalent would be like comparing a console’s power meter with the whole industry’s emissions ledger. The numbers provide context, not a fake trend.

Which Gaming Emissions Claims Deserve Trust?

A gaming-emissions claim deserves trust only when its boundary, data source, and assumptions are clear. A dramatic total without those details is just a big number wearing a lab coat.

• Boundary: Does the claim measure electricity use, life-cycle emissions, manufacturing, or a broader collection of gaming activities? Those are different questions, so they produce different answers.
• Evidence: Are the figures directly measured or estimated with proxies? Does the study disclose exclusions such as mobile gaming or physical-copy distribution?
• Assumptions: Which inputs drive the estimate, and how sensitive is the result to them?

Cloud gaming shows why that last check matters. The paper estimates about 6.92 million tonnes of CO₂-equivalent annually, using assumptions that include 400 million or more users, 4.5 hours of play per user each week, and 150 watts for data-center rendering hardware. The study lays out those inputs in its cloud-gaming methodology. The estimate is a modeled result, not a direct count of every cloud-gaming session.

We think the paper makes gaming’s meaningful modeled footprint harder to dismiss. But the exact total is still shaped by scope and assumptions. The headline is evidence worth taking seriously, not permission to turn estimates into certainty or blame individual players for an industry-wide model.

Ask What The Estimate Leaves Out

Your next step depends on what you want the estimate to answer. To understand your own gaming setup, make an inventory of your platform, how games reach you, your play habits, and whether your hardware is new, reused, or retired. That will show which parts of your situation the study cannot resolve. It will not produce a personal carbon score. Arithmetic without the missing inputs is just a spreadsheet in a funny hat.

If you’re citing the paper, open its methods and stated exclusions before repeating the headline. Keep the boundary beside the figure, and don’t present the modeled total as a complete count of gaming’s climate impact. A little context can stop a carefully qualified estimate from becoming an all-caps claim with a number attached.

If you’re comparing a specific activity, write down the question first: console electricity, device production, game delivery, or the broader industry footprint. Then check whether the study actually measures that thing. If it doesn’t, mark the answer as unknown instead of filling the gap with a guess. A blank in the ledger is more honest than a made-up boss fight.