Executive Summary
The F-22 Raptor is still the best air-superiority fighter ever built. It wins the one-on-one duel against China's J-20, and it loses the war anyway. That is not a contradiction — it is the entire intelligence argument.
One aircraft was built to dominate the final chapter of the Cold War, a European fight against Soviet Flankers that never came. The other was built for a Pacific war that hasn't started yet. The honest question over Kadena at 0300 is not whether the Raptor beats the J-20 in a dogfight. It is whether the Raptor reaches the fight at all, whether there are enough of them to matter, and whether the bases they stage through survive the opening hour.
Iron Command Assessment: In isolation — one Raptor against one J-20 — the Raptor wins on stealth, sensor fusion, pilot training and two decades of combat ecosystem. But the measures that decide a theatre are different: aircraft numbers, missile range, base survivability, and replacement timeline. On every one of those, the American position in 2026 is worse than it was in 2016. The F-22 does not lose to the J-20 in the sky. It loses to the clock — to the closed production line, the Chengdu factory floor, the missile-range gap, and every year Washington spends arguing about the aircraft meant to replace it. (Confidence: high on the structural gap; the duel-level superiority of the F-22 is not in dispute.)
Head-to-Head
| F-22 Raptor | J-20 | |
|---|---|---|
| First flight / service | 1997 / 2005 | 2011 / 2017 |
| Frontal radar cross-section | ~0.0001 m² (marble-sized, measured) | Larger; estimates only, not public |
| Engines | F119, supercruise ~Mach 1.5, thrust-vectoring | WS-15 indigenous (first production-colour flight Jan 2026) |
| Primary BVR missile | AIM-120D, ~160 km | PL-15 / PL-17, up to ~400 km |
| Two-seat / drone control | None in service | J-20S twin-seat, controls unmanned wingmen (since mid-2025) |
| Production | Line closed 2011 at 187 built | 100-120 per year; RUSI projects ~1,000 by 2030 |
| Combat-coded & available | ~57 at a 40% mission-capable rate | Growing fast |
| Replacement | F-47, mid-2030s | J-36 sixth-gen, three prototypes already flying |
1. The duel: the Raptor is genuinely the best
In 2005 Lockheed fielded an aircraft with a frontal radar cross-section measured at roughly one ten-thousandth of a square metre — the size of a marble, not as a metaphor but as a measurement. Twenty-one years on, nobody has matched it: not China, not Russia, not even the F-35. Pair that stealth with sustained Mach 1.5 supercruise without afterburner, and thrust-vectoring nozzles that let the jet point where its wings don't support, and against the threat it was designed for — Soviet Flankers over central Germany — the Raptor was invincible.
Through the 2010s Red Flag exercises, F-22 pilots killed aggressor aircraft at rates the Air Force won't publish in full, shooting F-16s and F-15s from beyond visual range before the aggressors knew a Raptor was airborne. If a Raptor finds your aircraft, you are already dead. None of that is in question. It is simply the wrong thing to measure.
2. The J-20 isn't trying to win the duel
In 2011 — the year the Raptor production line closed — a twin-tailed delta took off from Chengdu. The J-20 is not built to meet the F-22 in a turning fight. It is the instrument of an anti-access, area-denial doctrine: range, numbers, sensors and missiles designed to push America out of the first island chain before a Raptor ever fires.
The point of the J-20 is the missile it carries. The first close-up photographs of the PL-17 surfaced on 27 January 2026 in Chinese state-adjacent media, confirming what American analysts had suspected for two years: a roughly 400 km air-to-air missile, operational since 2023. The best missile in the American arsenal, the AIM-120D, reaches about 160 km. At that gap the PL-17 doesn't need to kill Raptors. It kills the tankers, the AWACS and the P-8s — the aircraft the Raptor depends on to reach the fight, find its targets, and stay airborne. Washington's answer, the AIM-260, has slipped its service entry four years running; Lockheed received a roughly $1 billion production contract to begin in September 2025, and as of early 2026 the weapon is still not operational.
So the Raptor can score a J-20 kill over the Taiwan Strait and still run out of fuel flying home, because the tanker it needed is burning on the apron at Kadena. The duel was won. The battle was already lost.
3. The 187-plane problem
China is building between 100 and 120 J-20s a year. The American F-22 line closed in 2011 at 187 airframes, full stop. RUSI projects around 1,000 J-20s in Chinese service by 2030, putting the Pacific ratio by the end of the decade at something like five to one or worse. That is not an aircraft gap. It is a category gap.
The numbers get worse on inspection. Of the 187 Raptors built, 143 are combat-coded; the other 32 are Block 20 training airframes Congress has forced the Air Force to keep flying. In FY2024 readiness data reported by Air & Space Forces Magazine, the mission-capable rate of the combat-coded fleet was 40%. That means at any given moment the US Air Force has roughly 57 F-22s able to fly a combat sortie — 57, to cover a Pacific theatre the size of the continental United States, Alaska and Mexico combined. That figure has been trending down for three years. It is not a bad quarter; it is the system tapping out. The Air Force sees it: the FY2026 budget requests around $90 million for a new infrared defensive system and fresh stealth coatings — buying the Raptor time it does not fully have.
4. The bases don't survive the opening hour
The cold open is not aimed at Raptors in the air. It is aimed at the airfields they stage through. A DF-26 with a 4,000-5,000 km range puts every forward airfield America operates in the first island chain — Kadena, Misawa, Elmendorf, Anderson on Guam — inside Chinese strike range. A Center for Strategic and International Studies war game run 24 times found the same outcome every time: around 90% of American aircraft destroyed on the ground in the opening hours. Fifty-seven Raptors staging through airfields the Chinese rocket force has already ranged is not a deterrent. It is a target list.
5. The engine gap and the replacement clock
The technical edge Washington spent two decades exploiting is closing. On 24 January 2026 the first J-20A with indigenous WS-15 engines flew in production colours. The J-20S twin-seat variant, operational since mid-2025, is the world's first stealth fighter designed to control unmanned wingmen from the cockpit — a capability no American fifth-generation aircraft has in operational service today.
And the replacement clock favours China. America's F-47, the intended F-22 successor, is on track for a first flight around 2028 but not expected in service until the mid-2030s on the Air Force's own revised estimate. China's sixth-generation effort is already in the air: a three-engine stealth prototype, provisionally designated J-36, flew from the Chengdu line on 26 December 2024, and by December 2025 a third prototype was airborne — three airframes flying inside twelve months. That is a roughly ten-year gap, during which the USAF has to deter China with an airframe whose design was finalised in 1997.
The verdict
F-22 wins the duel. J-20 wins the war.
The F-22 Raptor in 2026 is a deterrent against a war that has already moved past it. By 2028 it is likely to be a legacy aircraft with a fresh coat of paint, guarding airfields within range of weapons it was never designed to survive. It does not lose to the J-20 in the sky. It loses to the factory floor in Chengdu, to the missile-range gap, and to every year spent arguing about its successor. That is what a losing position in peer deterrence looks like — not a dog bite, but 57 Raptors and a 40% mission-capable rate.
Iron Command Assessment: The F-22-versus-J-20 question is a trap if you score it as a platform duel — that scoreboard favours America and tells you nothing useful. Score it on the four measures that decide a theatre — mass, missile reach, base survivability, and replacement timeline — and the American position has degraded on all four since 2016. The decisive matchup is no longer F-22 versus J-20; it is J-20 (and J-36) versus F-47, the fight that decides the 2030s. (Confidence: high on the trajectory. Revisit on F-47 first flight and the next published F-22 mission-capable rate.)
This piece is the written companion to the Iron Command comparison "F-22 vs J-20: America's 187-Plane Problem". For the full assessment with every figure graded, watch the video above or read it each week in the Pacific Brief.
Related assessments
Same methodology, different question
America's 11-Year Missile Gap: The PL-17 vs the AIM-260
China's PL-17 air-to-air missile has been operational since 2022. America's answer, the AIM-260, still isn't fielded — and was cleared for export before the USAF could use it. An operator's read of the Pacific air-superiority gap.
F-35A vs Su-57: A Comparative Assessment of Fifth-Generation Air Superiority
Military intelligence assessment comparing the F-35A Lightning II to Russia's Su-57 Felon. Stealth, sensors, combat record, and production — which fifth-gen fighter wins?
America Promised Australia Submarines It Can't Yet Build: The AUKUS Industrial Gap
AUKUS Pillar 1 commits the US to transferring Virginia-class submarines to Australia from the early 2030s. The binding constraint isn't money or political will — it's shipyard throughput. An operator's read of the gap, with a three-branch forecast.
Pacific Brief
Get the next comparison in your inbox
Written intelligence assessments with analyst notes. Mon / Wed / Fri.