Proven oil reserves would last about 55 years at today’s rate

If the world kept pumping oil at its 2024 rate, its proven reserves (the oil it knows is there and can profitably extract) would run out around 2080. That is BGR’s 252 billion tonnes of reserves divided by the 4.55 billion tonnes produced in 2024; BGR is Germany’s federal geoscience agency⁹.

That date is not a forecast. In 1980 the same sum gave about 30 years. Since then the world has burned more oil than the whole of the 1980 reserves, and the years left have still gone up¹⁷. The figure works like a fuel gauge that reads higher every time someone re-measures the tank. The US Energy Information Administration calls it “an often cited, but misleading, measurement”, because proven reserves count only known projects, and expects reserves to keep growing¹⁸.

Other official figures land close by. Using the International Energy Agency’s (IEA) reserves and production, the same sum gives about 49 years; using OPEC’s crude-only figures, about 58¹²¹¹. The differences come down largely to what each counts as oil: OPEC leaves out Canada’s oil sands, while BGR counts natural gas liquids (light liquids that come up with natural gas)¹¹¹⁰. On 2018 figures the same sum ran out in 2068; seven years later it runs out in 2080.

About 1.76 trillion barrels of oil are left that we know how to pump

The world has about 1,756 billion barrels of proven oil reserves, according to the IEA’s latest count, which it compiles from BGR, the Energy Institute and other sources¹². Counting crude oil alone and leaving out Canada’s oil sands, OPEC puts the figure at 1,572 billion barrels at the end of 2025¹¹.

Proven reserves are only part of what is in the ground. The IEA puts the oil that could be recovered with today’s technology at about 6,161 billion barrels, more than three times the proven reserves¹². The IEA finds there is enough to meet demand to 2050 in every one of its scenarios¹².

The world has pumped almost as much oil as it has left

All the oil ever pumped comes to about nine tenths of what is left in proven reserves: 223 billion tonnes against 252, by BGR’s count, which measures oil by weight⁹. Using up today’s reserves would mean burning slightly more oil in the coming decades than in the whole history of the industry so far.

BGR also counts about 498 billion tonnes of further resources: oil that is known or expected to be there but isn’t yet economic or proven⁹. Some of that will become reserves.

Oil reserves keep growing because they measure what is worth pumping

Proven reserves are an economic measure as well as a geological one. The industry standard defines them as oil that can be “estimated with reasonable certainty to be commercially recoverable” from known fields under current economic conditions, operating methods and regulations¹⁹. When prices, technology or rules change, so do the reserves.

The record shows it. In 1980 the world had about 683 billion barrels of proven reserves, and by 2020 it had 1,732 billion, even though it had pumped about 1,093 billion barrels in between, by our sum of the yearly figures. The years left rose from about 30 to 53.5¹⁷.

Price moves reserves quickly. When the US oil price fell 28% in 2020, US proven reserves fell 19%, because some oil was no longer worth pumping²⁰. Technology moves them more slowly but further: fracking in shale lifted US crude reserves from 19.1 billion barrels in 2008 to 44.2 billion in 2019, and shale now holds 60% of the country’s reserves²¹²².

Some of the biggest jumps came from counting oil that was already known. Between 2007 and 2010 Venezuela’s reserves tripled, from 99 billion to 297 billion barrels, largely by adding extra-heavy oil from its Orinoco Belt (oil too thick to flow without heating or diluting). That was about 78% of the world’s net addition in those years, by our arithmetic on OPEC’s tables²³. Canada’s oil sands, about 175 billion barrels of bitumen, entered the Energy Institute’s series in 1999 and the Oil & Gas Journal’s count in 2003¹⁷²⁴, but OPEC’s tables still count only Canada’s conventional crude²³.

Politics plays a part too. Researchers at University College London argue that some Middle Eastern figures include “political reserves”, booked “to convey a political message of resource abundancy”. On a strict industry definition they put the world’s proven reserves at about 1,276 billion barrels in 2018, against 1,700 to 1,790 billion in the IEA, bp and BGR figures²⁵.

Venezuela and Saudi Arabia hold the most oil

Venezuela has the largest proven crude oil reserves in the world, about 304 billion barrels, followed by Saudi Arabia with 267 billion, according to OPEC¹¹. OPEC’s members together hold 79% of the world’s proven crude. Some of those figures haven’t changed in years: Iran and Kuwait reported exactly the same reserves every year from 2021 to 2025¹¹.

Venezuela pumps little of its oil. It has about 17% of the world’s crude reserves, mostly extra-heavy Orinoco oil that is costly to extract, but produced only 0.8% of the world’s crude in 2023³.

For the countries that live on oil money, big reserves cut both ways. If the world stops buying oil before they have pumped it, some of it will never be sold. The Intergovernmental Panel on Climate Change (IPCC) warns that limiting warming to 2 °C or below “will leave a substantial amount of fossil fuels unburned”, and puts the value of the unburned fuels and stranded infrastructure (plants and fields made worthless before they are used up) at around $1 trillion to $4 trillion¹. A shrinking market may still favour the biggest holders: on the IEA’s 2021 path to net zero, OPEC’s share of a much smaller oil supply rises from about 37% to 52%².

A 2020 IMF staff paper found that the Gulf oil states, which pump a fifth of the world’s oil, could run through their financial wealth by 2034 at their spending of the time, though it predates the high oil prices of 2022⁴.

Most of the oil can’t be burned if warming is to stay near 1.5 °C

The oil already in proven reserves is more than a 1.5 °C world can take. Burning all 1,756 billion barrels would release roughly 750 billion tonnes of carbon dioxide, at about 0.43 tonnes a barrel, by our arithmetic¹²²⁹. That is an upper bound, since some oil ends up in plastics and asphalt rather than being burned.

From the start of 2026, the world could emit only about 130 billion tonnes of CO2 for an even chance of holding warming to 1.5 °C, what scientists call the remaining carbon budget, and about 1,050 billion tonnes for 2 °C²⁸. At 2025’s emissions the 1.5 °C budget would be gone in a little over three years²⁸.

The oil reserves alone would also use up nearly three quarters of the 2 °C budget. A 2021 study in Nature found that nearly 60% of the world’s oil reserves would have to stay unextracted in 2050 for an even chance of 1.5 °C, with oil and gas production falling 3% a year²⁶²⁵. A 2015 study aiming at 2 °C put the share at a third²⁷. The IEA’s 2021 path to net zero emissions needed no new oil and gas fields at all².

Oil demand may peak before 2030, or not for decades

Whether oil demand peaks soon depends on whom you ask. In the IEA’s main long-term scenario, which assumes the policies governments have announced, demand peaks around 2030 at 102 million barrels a day¹². Its shorter-term market report, counting differently, has demand levelling off near 105.5 million barrels a day by 2030, oil burned as fuel peaking as early as 2027, and China’s demand on track to peak this decade¹⁴.

OPEC sees no peak at all, and says so with some insistence.

OPEC expects total energy demand to grow by 23% to 2050, almost entirely in developing countries¹⁶. It also points to its record: its first outlook, in 2007, expected around 110 million barrels a day for 2025, and actual demand came in a little over 105¹⁶. OPEC presents that as close; it was also about 5% too high. And if today’s policies simply stay as they are, the IEA’s own scenario has demand still rising in 2050, to 113 million barrels a day¹².

In 2026, meanwhile, the Iran war has restricted supply, and the IEA expects world demand to fall by 2.6 million barrels a day¹³. A fall forced by scarce, expensive oil probably says little about whether demand has peaked.

Electric vehicles are already cutting oil use

Electric vehicles of every kind, from scooters to buses, saved about 1.7 million barrels of oil a day in 2025⁶. That is about one barrel in sixty of what the world used, by our arithmetic¹⁴. Norway shows how far it can go. About 97% of new cars sold there in 2025 were electric, and more than a third of all the cars on its roads⁶. Deliveries of petrol and road diesel have fallen by about a quarter since 2016, by our arithmetic on Statistics Norway’s figures⁷.

The gap between those numbers suggests how slowly a country’s cars turn over: even in Norway, most cars on the road still burn fuel. How fast the rest of the world is following is on the page about the EV tipping point.

The world has also been getting more out of each barrel for half a century. In 1973 it took a little under one barrel of oil to produce $1,000 of economic output; by 2019 it took 0.43, a fall of 56%⁸. Where oil is replaced outright, as in Norway’s cars, electricity takes its place, and a growing share of that comes from wind, sun and water, which all draw in the end on the sun’s energy.

How the age of oil could end

The 55 years will probably keep slipping, as the figure has for 45 years. The IEA finds enough oil underground to meet demand to 2050 in every scenario it runs, so the wells are unlikely to run dry first¹².

Plastics are the strongest argument that oil demand will last. By 2030 making polymers and synthetic fibres will take more than one in every six barrels, a use that electric cars do nothing to reduce¹⁴.

The end of oil is more likely to come from the demand side of the sum. To keep warming near 1.5 °C, the IPCC’s pathways cut oil use by about 60% by 2050 compared with 2019¹, close to the 3% a year decline the Nature study found necessary²⁶. In the IEA’s main scenario demand is close to its peak; in OPEC’s, and in the IEA’s own scenario with no new policies, it is nowhere in sight. Which is right depends on how fast electric vehicles spread, how much plastic the world makes and what governments do about the carbon budget, far more than on what is left underground.