Who actually runs the world's bike share fleets

Any ranking of bike share operators by fleet size is a ranking of press releases. The operators with the largest fleets publish the least about them, and the operators that publish most are the ones a public authority made publish.

So this list is ordered by evidence rather than by scale. It runs from the system whose economics an independent public body has forced into daylight, down to the operators about whom nothing checkable exists at all.

1. Lyft, via Citi Bike in New York

Best documented, and not by choice.

New York City’s Independent Budget Office published Citi Bike: Lessons for the Future of New York City’s Bike Share in November 2025, ahead of a contract that expires in May 2029. It records over 44 million rides on approximately 37,000 bicycles in 2024, making Citi Bike the largest bike share system in the United States, with ridership more than doubled since 2019.

The financial findings are the useful part. Citi Bike generated $200 million in revenue in 2024. Electric bikes accounted for 66% of all rides, and revenue from per-minute charges on those rides represented 40% of total revenue. An annual membership rose from $95 in 2013 to $220 in 2025.

Then the sentence that defines this entire sector. “Detailed information about the program’s operating costs is not publicly available.” The budget office had to estimate them, putting hard costs such as bicycles, docks, and the personnel and vehicles needed to rebalance and recharge at about $128 million, and noting that the remaining $72 million covers some mix of executive pay, advertising, insurance and profit.

That is a city with a signed contract, a transportation department, a comptroller and a statutory budget office, and it still cannot see the cost base of its own bike share. One further finding is worth carrying: Citi Bike achieves 3.3 trips per bike per day, the highest of the American systems reviewed, and both Montréal and Paris do better.

2. Serco, via Santander Cycles in London

Second best documented, because Transport for London is a public body that announces its contracts.

TfL announced on 11 December 2025 that Serco would operate the scheme until 2031, with an option to extend by five years. The same release reports 9 million hires in the twelve months to October 2025, a 4% increase, 109 million hires since 2015, a record 728,000 member hires in October 2025, and 2,000 electric bikes in the fleet.

Nine million hires a year is a useful corrective. London’s scheme is among the most photographed in Europe and it runs at roughly a fifth of New York’s volume. Size in this sector is not correlated with visibility, and it is certainly not correlated with how often a system appears in a slide about urban mobility.

What TfL does not publish alongside that is the contract value, which means the cost per hire, the single most informative number about any bike share, remains as unavailable in London as it is in New York.

3. Lime

The largest operator that reports voluntarily, and the limits of voluntary reporting are visible in the reporting itself.

Lime’s account of its 2023 results, published on 8 April 2024, reports 156 million trips, gross bookings of $616 million and adjusted EBITDA above $90 million, with monthly active riders averaging 3.1 million.

Read the categories rather than the numbers. Trips, not bike trips. The company operates bicycles and standing scooters and reports them together, and no figure in the release separates the two. There is no fleet size. There is no audited financial statement behind any of it, because Lime is private and publishes by announcement rather than by filing.

None of that makes the figures false. It makes them unverifiable, which is a different problem and in some ways a worse one, because unverifiable numbers get quoted with the same confidence as audited ones.

4. Meituan

The operator most often assumed to be the largest, and the one about which least can be established.

Meituan is listed in Hong Kong and files audited accounts, so the disclosure exists. Its announcement of results for the year ended 31 December 2025 reports two segments. Core Local Commerce turned over RMB 260.8 billion. New Initiatives, which is where bike sharing sits, turned over RMB 104.0 billion with an operating loss of RMB 10.1 billion.

That is close to the entire public record. The business review and management discussion in that announcement run to several pages and do not describe the bicycle business at all. No trip count, no fleet size, no revenue line, no city count.

Bicycles appear once, in the glossary defining what the group counts as a transaction, and the definition is instructive. Where a rider holds a monthly pass, only the purchase or claim of the pass is a transaction and the rides taken on it are not. So even the group’s headline transaction figure, if anyone tried to use it as a proxy, would not be a count of journeys.

Which means the operator most often placed at the top of a global ranking cannot be placed on one from anything it has published, and neither can it be ruled out. Every figure in circulation purporting to size China’s bike share market has been reconstructed from outside, and no ranking in this article, including the order of this list, should be read as a claim about who runs the most bicycles.

5. Everyone else

Hellobike and the other large Chinese operators are private and publish nothing an outsider can check. European operators are mostly contractors delivering municipal schemes, and their figures live inside city contracts that are published unevenly and rarely in a comparable form.

The one number that travels

If a reader takes a single metric from this, make it trips per bike per day.

It is the only figure in the sector that compares across systems of different sizes, ownership models and currencies. Fleet size flatters whoever has the most bicycles parked. Total rides flatters whoever operates in the biggest city. Revenue flatters whoever charges most. Trips per bike per day measures whether the machines are being used, which is the thing every other number is a proxy for, and it is the number an operator with idle stock least wants published.

New York’s 3.3 trips per bike per day comes with the useful detail that two other cities beat it, which tells a procuring authority that the American benchmark is not the ceiling. Neither London nor any of the private operators publishes the equivalent, and in London’s case the components are held by a public body that could publish it tomorrow.

What a procuring authority should take from this

The ranking above is really a ranking of contracts. Where a public body wrote disclosure into an agreement, or where an independent budget office had standing to ask, there are numbers. Everywhere else there are announcements.

New York’s experience contains the specific warning. Its budget office notes that Lyft owns the system’s infrastructure, and that a different operator would have to provide new equipment or purchase the existing estate. A city that does not own the docks has not bought a service; it has rented a dependency, and the price of leaving is set by the incumbent.

Any authority letting a scheme should therefore treat cost transparency, asset ownership at expiry and a published utilisation measure as contract terms rather than as things to request later. They are cheap to insert before signature and impossible to obtain afterwards.

The wider pattern, in which the adoption of a technology is settled by who is paying and on what terms rather than by the technology itself, runs through Norway’s electric car market and through the rest of this desk’s work in emerging technology.