Missions and launches
How many satellites orbit the Earth?
The current figures say one thing clearly: the problem in low Earth orbit is not the satellites that work, it is the pieces left behind when they stop.
In brief
There are around 16,000 still-working satellites around Earth, out of more than 18,800 left in orbit after seventy years of launches, and surveillance systems keep about 46,200 objects in their catalogue. They almost never collide because the available space is enormous, the orbits are known, and satellite operators receive close-approach warnings and move their spacecraft.
Key points
- About 16,000 active satellites, 18,840 objects still in orbit from launches, 46,210 objects tracked by surveillance catalogues: the figures published by the European space debris office as of late July 2026.
- Since 1957 there have been roughly 7,320 successful launches, which put about 27,490 satellites into orbit.
- The total mass of everything in orbit exceeds 17,000 tonnes.
- Fragments are the real problem: about 54,000 objects above 10 centimetres and 1.2 million between 1 and 10 centimetres, mostly too small to be tracked individually.
- A collision between two satellites has already happened, on 10 February 2009, and produced more than 2,300 trackable fragments.
The short answer is: about 16,000 working satellites. But it is one of those figures that says little on its own, because the interesting number is a different one, and it is larger.
Today’s numbers
The space debris office of the European space agency, in Darmstadt, maintains a page of statistics on the orbital environment that is updated continuously. At the most recent update, in late July 2026, the picture is this:
- about 7,320 successful launches since the start of the space age in 1957;
- about 27,490 satellites placed in orbit by those launches;
- about 18,840 still physically in orbit;
- about 16,000 still working;
- about 46,210 objects regularly tracked by surveillance networks and kept in the catalogue;
- more than 660 fragmentation events, counting break-ups, explosions and collisions;
- more than 17,000 tonnes of total mass.
The gap to read is the one between 16,000 and 46,210. The difference is not satellites: it is spent upper stages, instrument covers, launch adapters and above all fragments. And the catalogue only sees what is big enough: below 10 centimetres in low orbit most objects escape tracking and are estimated with statistical models. Those estimates point to about 54,000 objects above 10 centimetres, 1.2 million between 1 and 10 centimetres, and 140 million between 1 millimetre and 1 centimetre.
Why they don’t collide
Three reasons, in order of importance.
Orbital space is enormous. The visual intuition from illustrations, with Earth wrapped in a dense cloud of dots, is misleading: those dots are magnified millions of times. The shell between 200 and 2,000 kilometres of altitude has a volume on the order of a thousand billion cubic kilometres. Sixteen thousand objects scattered through that volume are, on average, extraordinarily isolated.
Orbits are not random. A satellite does not wander: it follows a calculated trajectory, and at certain altitudes the slots are assigned. Geostationary orbit, at 35,786 kilometres, is a single precious ring because it is the only one where a satellite stays fixed above the same point on the ground: positions on that ring are coordinated internationally.
Operators get warnings and move their spacecraft. The positions of catalogued objects are propagated into the future and, when two trajectories come too close, operators receive a conjunction warning. At that point they decide whether to fire thrusters for an avoidance manoeuvre, typically a few centimetres per second of velocity change, which applied a few hours ahead shifts the pass by kilometres. It is routine for control centres, not an emergency.
The times it went wrong
On 10 February 2009, at 16:56 UTC, at 776 kilometres above Siberia, the American communications satellite Iridium 33 and the defunct Russian military satellite Kosmos 2251 collided at 11.7 kilometres per second. It was the first accidental collision between two satellites in history. Both were destroyed and the impact produced more than 2,300 trackable fragments, many of which are still up there. The details are in the European space debris briefing.
Two years earlier, in January 2007, the deliberate destruction of the Chinese weather satellite Fengyun 1C during an anti-satellite test increased the population of trackable objects by 25 per cent on its own. A single event, a quarter of the catalogue.
From here you can see why speed matters more than mass. A one-centimetre aluminium fragment weighs about a gram and a half, but at 10 kilometres per second it carries energy comparable to a small car travelling at fifty kilometres an hour. Against a pressurised wall that is not a scratch.
The risk the field calls by name
The scenario that worries people has a precise name, the Kessler syndrome: every collision generates fragments, fragments raise the probability of further collisions, and past a certain density the process feeds itself even without new launches, until the population is ground down into pieces too small to be dangerous. This is not a film plot: it is the reason mitigation rules exist, such as venting the tanks of spent stages so they cannot explode, and designing satellites to re-enter and burn up in the atmosphere within a few years of the end of service.
It is worth remembering that low orbit is not an eternal parking space: at those altitudes traces of atmosphere remain, and drag brings objects down over time. It is the same reason the Space Station has to be boosted periodically, and why staying in orbit is about speed, not altitude. Low orbit cleans itself, but on its own schedule: years for the lowest debris, centuries or millennia for the higher pieces.