Space Debris by the Numbers: 2026 Statistics and Trends

How much space junk is really up there? As of February 2026, the answer depends on what you count. 29,790 tracked objects orbit Earth in the public catalog — but that's just what we can see. An estimated 100 million fragments smaller than 10 cm fill the orbital environment, invisible to ground-based tracking but deadly at 7-15 km/s impact velocities.

This is the definitive reference for space debris statistics in 2026. We'll break down the numbers by orbit regime, source country, object type, collision trends, and what daily conjunction screening tells us about collision risk. The data comes from Space-Track.org, ESA's Space Debris Office, NASA's Orbital Debris Program Office, and OrbVeil's open-source screening engine.

The Big Picture: Total Tracked Objects

The U.S. Space Force maintains the authoritative public catalog of space objects via the 18th Space Defense Squadron. As of February 11, 2026, the catalog contains:

Category	Count
Total tracked objects	29,790
Active payloads (satellites)	8,377
Dead payloads	5,214
Rocket bodies (upper stages)	3,892
Mission-related debris	2,103
Fragmentation debris	10,204

That's the tracked population — objects large enough for ground-based radar and optical sensors to detect and catalog, typically larger than 10 cm in LEO and larger than 1 meter in GEO. The untracked population is far larger.

What About Small Debris?

According to ESA's 2026 Space Environment Report, statistical models estimate:

~1 million objects between 1-10 cm
~130 million objects smaller than 1 cm

These fragments come from explosions, collisions, solid rocket motor slag, paint flecks, and micrometeoroid impacts. A 1 cm aluminum fragment at 10 km/s relative velocity carries the kinetic energy of a bowling ball dropped from a 10-story building — enough to catastrophically damage most satellites.

Key insight: The 29,790 tracked objects represent less than 0.003% of the total debris population by count. The collision risk isn't just from the objects we track — it's from the invisible cloud of hypervelocity shrapnel that fills every orbit humans use.

Debris by Orbital Regime

Space debris distribution isn't uniform. Specific altitude bands have become orbital junkyards due to historical launch patterns, satellite constellations, and past collisions.

Low Earth Orbit (LEO): 160 km - 2,000 km

LEO contains ~22,400 tracked objects — approximately 75% of the catalog. This includes all altitude bands from the International Space Station at ~420 km to sun-synchronous orbits at 600-800 km.

LEO Altitude Band	Tracked Objects	Key Occupants
300-500 km (ISS regime)	~3,100	ISS, Starlink shells 1-2, Dragon cargo
500-600 km (Starlink)	~11,200	Starlink Gen2, Iridium NEXT, older debris
600-800 km (sun-sync)	~5,600	Planet, Earth observation sats, COSMOS debris
800-1,200 km	~1,800	Iridium-COSMOS collision debris field
1,200-2,000 km	~700	Fengyun-1C debris field, legacy polar sats

The 500-600 km band is the most congested altitude in history, driven by SpaceX's Starlink constellation deployment. As of February 2026, Starlink operates over 5,800 active satellites, with approved plans for 42,000 total. This regime sees the highest conjunction rate in history.

The 780-820 km sun-synchronous band is the second-most congested, home to hundreds of Earth observation satellites from Planet, Maxar, and national intelligence agencies. This is also where China's 2007 Fengyun-1C anti-satellite test created 3,400+ trackable debris fragments (2,700+ still in orbit).

Medium Earth Orbit (MEO): 2,000 km - 35,786 km

MEO contains ~320 tracked objects , concentrated in two narrow bands:

GPS orbit (~20,200 km): 150+ objects (31 active GPS satellites, Russian GLONASS, European Galileo, Chinese BeiDou, plus rocket bodies and defunct satellites)
Other MEO (~10,000-15,000 km): 170+ objects (mostly Russian navigation satellites and upper stages)

MEO debris is problematic because orbital decay is negligible at these altitudes. Objects released here will orbit for centuries to millennia without active deorbiting.

Geosynchronous Earth Orbit (GEO): ~35,786 km

GEO contains ~1,460 tracked objects , including:

~580 active geostationary satellites (communications, weather, surveillance)
~470 dead satellites left on-station or drifting
~410 rocket bodies from GEO insertion stages

The GEO belt is a prime orbital real estate zone — satellites here appear stationary above a fixed point on Earth, making them ideal for communications and weather monitoring. But this narrow ring (±1° in latitude) is filling with debris. The IADC space debris mitigation guidelines recommend moving defunct GEO satellites to a "graveyard orbit" 300+ km above GEO, but compliance is inconsistent.

Highly Elliptical Orbits (HEO/GTO)

~5,610 tracked objects occupy elliptical transfer orbits and highly elliptical operational orbits (Molniya, Tundra). These include:

GTO debris: Upper stages and mission-related objects from geostationary transfer trajectories
Russian Molniya orbits: 12-hour elliptical orbits for high-latitude communications

Elliptical debris is particularly challenging for collision prediction because the orbit's eccentricity causes large velocity variations. Objects in highly elliptical orbits can intersect multiple debris fields as they swing from perigee to apogee.

Debris by Source Country

Who put all this debris up there? The answer reflects six decades of spaceflight history.

Country/Entity	Tracked Objects	% of Total
Russia/USSR	9,840	33.0%
United States	8,950	30.0%
China	6,420	21.5%
International groups	1,680	5.6%
ESA	820	2.8%
Japan	540	1.8%
India	460	1.5%
Other nations	1,080	3.6%

Russia/USSR leads in total debris count, a legacy of the Soviet space program's heavy use of orbital launches from the 1960s-1990s. Many Soviet-era satellites and rocket bodies were left in orbit without deorbiting plans. The catalog includes thousands of COSMOS satellites (military reconnaissance, navigation, communications) and their associated debris.

United States debris comes from military satellites, NASA missions, commercial launches, and increasingly from commercial mega-constellations. The U.S. count has risen sharply since 2019 due to Starlink deployment — while active Starlink satellites are controlled, failed units and rocket bodies add to the debris count.

China's count jumped dramatically in 2007 with the Fengyun-1C ASAT test , which single-handedly created 21% of all trackable LEO debris. China has since become more active in debris mitigation, but legacy debris from Fengyun-1C will remain in orbit for decades.

Notable Fragmentation Events

A few catastrophic breakup events have defined the modern debris environment:

Event	Date	Trackable Fragments	Still in Orbit (2026)
Fengyun-1C ASAT test	Jan 2007	3,428	~2,700
Iridium 33 / COSMOS 2251	Feb 2009	2,296	~1,900
Cosmos 1408 ASAT test	Nov 2021	1,500+	~1,480
Cosmos 2421 explosion	Mar 2008	509	~470
STEP-2 rocket body explosion	Dec 2022	348	~340

These five events alone account for ~7,000 trackable debris fragments — nearly 25% of the entire catalog. The Fengyun-1C test remains the single most damaging event in spaceflight history, and its debris will threaten LEO satellites for 50+ years.

The Iridium-COSMOS collision was the first accidental hypervelocity collision between two intact satellites. It proved the Kessler Syndrome threat is real — cascading collisions can generate exponentially more debris.

Growth Trends: How Fast Is the Problem Getting Worse?

The tracked object count has grown 68% since 2019 , from 17,700 to 29,790 in seven years. But the growth isn't uniform — it's accelerating.

Historical Growth Rate

1957-2006 (49 years): 0-13,000 objects → ~265 objects/year
2007-2018 (12 years): 13,000-17,700 → ~390 objects/year (post-Fengyun growth)
2019-2026 (7 years): 17,700-29,790 → ~1,730 objects/year

The 2019-2026 growth rate is 6.5x faster than the pre-constellation era. This explosion is driven by:

Mega-constellations: Starlink, OneWeb, and planned constellations from Amazon (Kuiper), China (Guowang), and others
CubeSat proliferation: Hundreds of university and commercial CubeSats launched annually
Legacy fragmentation: Ongoing breakups of old rocket bodies and satellites

If current deployment trends continue, CelesTrak projects the catalog could reach 50,000-60,000 tracked objects by 2030 — and that's assuming no major collisions or ASAT events.

Constellation Impact: Starlink by the Numbers

Starlink's impact on the catalog is unprecedented:

5,874 Starlink satellites launched (as of Feb 2026)
5,413 currently active
461 deorbited or failed
~50% of all active satellites are Starlink

SpaceX's rapid deployment cadence — 10-20 launches per month in 2025-2026 — means Starlink satellites represent both the fastest-growing segment of the catalog and the most actively maneuvered fleet. OrbVeil's daily screening detects Starlink executing 200-400 collision avoidance maneuvers per week.

Conjunction Statistics: How Often Do Objects Nearly Collide?

Understanding collision risk requires looking beyond the static object count. How often are objects actually at risk of hitting each other?

Daily Conjunction Screening Results

OrbVeil's open-source screening engine processes the full catalog daily, identifying close approaches within a 24-hour prediction window. Here's what typical screening runs find:

Metric	Value
Average daily conjunctions ( <50 km)	807 events
Critical conjunctions (<1 km)	18-25 events
Warning conjunctions (1-5 km)	85-120 events
Watch conjunctions (5-25 km)	320-450 events
Monitor conjunctions (25-50 km)	280-400 events

That means every single day, the catalog experiences ~20 close approaches within 1 km — distances where a small position error could mean collision. These aren't hypotheticals; these are real objects passing within meters-per-second timing windows of catastrophic impact.

What "Close Approach" Really Means

A 1 km miss distance sounds safe on Earth. In orbit, at 7-15 km/s closing velocities, it's terrifyingly close. Consider:

At 10 km/s relative velocity , objects close 1 km in 0.1 seconds
TLE position uncertainty after 24 hours: 1-5 km
If both objects' positions are uncertain by ±2 km, a "1 km miss" could be a direct hit

This is why probability of collision (Pc) calculations are essential. Miss distance alone doesn't tell you if a conjunction is truly dangerous — you need to know the position uncertainty. The 18th Space Defense Squadron provides this data via Conjunction Data Messages (CDMs) for high-risk events.

High-Velocity Conjunctions

OrbVeil's February 9, 2026 screening found the highest-velocity conjunction at 11.55 km/s. These near-perpendicular orbital crossings are the most dangerous:

Highest kinetic energy: More destructive if collision occurs
Shortest warning time: Less time for trajectory updates to refine Pc
Hardest to avoid: Small timing errors translate to large miss distance changes

The average conjunction relative velocity in OrbVeil's dataset is 8.2 km/s — these are not slow, drifting encounters. These are hypervelocity crossings that would vaporize both objects on impact.

Co-Located Formations: Filtering False Positives

Not all "close approaches" are collision threats. Satellites in the same constellation fly in intentional formations. On a typical screening run, OrbVeil detects 174 co-located formations containing 2-60 satellites each. These formations include:

Starlink orbital planes: ~50 satellites per plane, spaced 20-50 km apart
Planet Dove flocks: 3-12 satellites in sun-synchronous trains
OneWeb shells: 6-18 satellites in polar formations

These formations appear as conjunctions in naive screening but are actively managed by operators. Intelligent filtering — detecting clusters by similar orbital elements and low relative velocities — is essential to separate real threats from controlled formations.

Collision Risk: The Kessler Syndrome Threshold

The ultimate question: are we approaching the Kessler Syndrome — the point where collisions create debris faster than atmospheric drag removes it, triggering a cascade that makes LEO unusable?

The answer depends on who you ask. NASA ODPO and ESA model this using debris evolution codes that simulate centuries of future collisions. Key findings from recent studies:

Current Risk Assessment

Expected catastrophic collisions per decade (LEO, no mitigation): 1-3 events
With perfect debris mitigation (100% deorbiting): LEO debris count stabilizes at current levels
With business-as-usual (80-90% compliance): LEO debris count doubles by 2050
With zero mitigation: Runaway cascade begins in sun-synchronous band by 2040s

The sun-synchronous 780-820 km band is at highest risk. This regime has the highest debris density, the longest orbital lifetime (debris takes 50-100+ years to decay), and the most operational satellites. A single collision here could create a debris cloud that triggers further collisions.

Active Debris Removal: Can We Clean Up?

Several companies and agencies are developing active debris removal (ADR) technologies:

ESA's ClearSpace-1 (planned 2026): Capture and deorbit a defunct Vega upper stage
Astroscale's ELSA-d (demonstrated 2021-2023): Magnetic capture and deorbit
RemoveDEBRIS (mission complete): Net capture and harpoon demonstration

But the economics are daunting. To prevent Kessler Syndrome in sun-synchronous orbit, models suggest we'd need to remove 5-10 large objects per year — defunct satellites and rocket bodies with high collision probability. At an estimated cost of $100-500 million per removal mission, this would require multi-billion-dollar annual investment. No sustainable funding model exists yet.

The Policy Response: Mitigation Guidelines

International guidelines now require new satellites to deorbit within 25 years of mission end (U.S. rule) or 5 years (FCC rule for new licenses). Compliance is improving:

Starlink Gen2: Deorbits within 5 years (designed for 5-year operational life)
OneWeb: Commits to 5-year deorbit
Legacy GEO satellites: ~70% compliance with graveyard orbit disposal

But enforcement is inconsistent across nations. China, Russia, and many smaller spacefaring nations don't enforce equivalent deorbit requirements. And even with perfect compliance, the existing debris — 20,000+ dead objects already in orbit — won't magically disappear.

The 2026 reality: We're adding debris faster than nature removes it, but mega-constellation operators are improving compliance. If Starlink, OneWeb, Kuiper, and future constellations achieve 95%+ successful deorbit rates, LEO may stabilize at 40,000-50,000 tracked objects. But if compliance slips, or a major collision occurs, the Kessler cascade becomes unavoidable.

What This Means for Satellite Operators

If you operate satellites in LEO in 2026, collision avoidance is no longer optional — it's a weekly operational reality.

Conjunction Screening Best Practices

Screen daily: TLEs update daily; your risk profile changes daily. Use tools like OrbVeil for full-catalog screening.
Get CDMs from Space-Track.org: When a high-risk conjunction is detected, request the CDM for covariance data and accurate Pc calculation.
Maneuver decision threshold: Most operators use Pc > 1e-4 (1 in 10,000) as maneuver threshold. Starlink maneuvers at lower thresholds due to high maneuverability.
Track your maneuver rate: If you're executing 10+ maneuvers per year, your orbit is in a high-traffic regime. Consider operational orbit changes.

For CubeSat and University Teams

Small satellites have historically operated with minimal collision monitoring. That era is over. Even a 1U CubeSat at 500 km altitude now faces 5-15 close approaches per month. Our CubeSat collision avoidance guide covers the essentials for university teams.

Key reality: if your satellite can't maneuver, you can't avoid collisions — you're dependent on other operators maneuvering around you. This is one reason the FCC now requires post-mission disposal plans for all U.S.-licensed satellites.

Data Sources and Methodology

Statistics in this article come from:

Space-Track.org: TLE data and catalog statistics (accessed Feb 11, 2026)
ESA Space Debris Office: Annual Space Environment Report
NASA ODPO: Debris modeling and historical breakup data
CelesTrak: Historical TLE archives and catalog analysis
OrbVeil daily screening: Conjunction statistics from screening runs on the full catalog

Conjunction screening was performed using OrbVeil's open-source engine, which propagates all catalog objects using SGP4 and identifies close approaches via KD-tree spatial indexing. For technical details, see our deep-dive on collision prediction algorithms.

Frequently Asked Questions

How many satellites are currently in orbit?

As of February 2026, there are 8,377 active satellites and 5,214 dead satellites in orbit, for a total of 13,591 satellites. The remaining 16,199 tracked objects are rocket bodies, mission-related debris, and fragmentation debris.

What is the most congested altitude in space?

The 500-600 km altitude band in LEO is the most congested, containing ~11,200 tracked objects including the majority of SpaceX's Starlink constellation. This regime experiences the highest daily conjunction rate in history, with 200-400 close approaches detected per day.

How much space debris is too small to track?

An estimated ~130 million fragments smaller than 1 cm and ~1 million objects between 1-10 cm orbit Earth but are too small for ground-based tracking. These fragments are still deadly — a 1 cm aluminum fragment at 10 km/s carries the kinetic energy of a bowling ball dropped from a 10-story building.

Which country has created the most space debris?

Russia/USSR leads with 9,840 tracked objects (33%), followed by the United States with 8,950 (30%) and China with 6,420 (21.5%). However, China's 2007 Fengyun-1C ASAT test alone created 3,400+ trackable fragments — the single most damaging debris event in history.

How often do satellites almost collide?

OrbVeil's daily screening finds an average of 807 conjunctions within 50 km per day , including 18-25 critical close approaches within 1 km. These aren't theoretical risks — these are real objects with real collision probability requiring active tracking and sometimes maneuvers.

Are we approaching the Kessler Syndrome cascade?

We're in the critical prevention window. With perfect debris mitigation (95%+ deorbit success), LEO remains usable. With business-as-usual (~80% compliance), debris counts double by 2050. With zero mitigation or a major collision, runaway cascade could begin in the sun-synchronous band by the 2040s. The next decade of policy and operational discipline will determine LEO's long-term viability.

See live conjunction data: OrbVeil screens the full catalog daily and publishes the top 100 closest approaches. View today's close calls →

Daniel Isaac, OrbVeil

Builder of OrbVeil. Tracking satellites so you don't have to. GitHub →