NASA tested a nuclear rocket engine dozens of times in the 1960s, proved it could work, then shelved it before it ever flew — and now the same basic idea is being revived because it could make the trip to Mars far shorter than chemical rockets allow.

In the 1960s, NASA and the Atomic Energy Commission built nuclear rocket engines and test-fired them dozens of times in the Nevada desert. The engines worked. Then the program was shelved before a single one ever flew. In recent years the same basic idea came back, driven by the promise that a nuclear engine could shorten the journey to Mars in a way chemical rockets cannot. The revival’s own story, it turns out, has been almost as stop-start as the original.

The physics is the constant in all this. The politics and the budgets are what keep changing.

What the 1960s actually built

The effort was called Project Rover, begun at Los Alamos and continued as a civilian program after NASA was formed. Its engine development strand became NERVA, short for Nuclear Engine for Rocket Vehicle Application. Between 1959 and 1972, the program ran 23 reactor tests at Jackass Flats, part of the Nevada Test Site, steadily working through designs that ran hotter, longer and at higher power.

By the end, the engines had been started, restarted and held at full thrust, and the technology was regarded as essentially proven on the ground. What it never got was a flight. A project to launch a reactor for an in-space test was cancelled, and Rover and NERVA themselves were shut down in early 1973. The reason was not that the engine failed. It was that the missions it was built for, crewed flights to Mars and permanent bases beyond the Moon, were never funded. With Apollo winding down and the Space Shuttle chosen as NASA’s next step, a powerful engine was left with nothing to push.

How a nuclear rocket works

A nuclear thermal rocket is not a bomb, and it does not explode. It uses the heat from a controlled nuclear reactor to do the same job a chemical rocket does with burning fuel. Hydrogen is pumped through the hot reactor, heated to extreme temperatures, and then blasted out of a nozzle to produce thrust.

The advantage is efficiency. A measure called specific impulse captures how much push you get from a given amount of propellant, and a nuclear thermal engine roughly doubles it, reaching around 850 to 900 seconds against the best chemical engines’ 450 or so. That does not mean twice the raw power. It means far more done with the same mass of propellant, which is exactly the currency that matters once you are trying to leave Earth far behind.

Why it appeals for Mars

For a trip to Mars, that efficiency translates into options. You can carry less propellant for the same manoeuvre, or spend the same propellant to travel faster, trimming the transit time. A shorter crossing is not just convenient. It means less time for astronauts to accumulate radiation exposure, less time in weightlessness, and fewer supplies to haul for the journey.

It is worth being careful with the phrase “far shorter.” A nuclear engine could meaningfully cut Mars transit times, but the exact saving depends heavily on how a mission is designed, and estimates vary. The honest version is that it improves the odds and widens the choices, rather than magically halving every trip.

Revived, then shelved again

The modern revival was real. Over the past several years, DARPA and NASA teamed up on a project called DRACO, the Demonstration Rocket for Agile Cislunar Operations, to build and fly a nuclear thermal engine and prove the idea in space at last. Lockheed Martin was chosen to develop the system, with a reactor designed to run on low-enriched uranium rather than the weapons-grade fuel of the 1960s, and a demonstration flight was targeted for the second half of the decade.

Then history rhymed. In 2025, DARPA cancelled DRACO, and the following budget confirmed the decision, leaving no funding for nuclear thermal or electric propulsion. The stated reason was not a technical dead end but a shift in the economics. The sharp fall in launch costs, driven by reusable rockets and the prospect of very large vehicles like Starship, weakened the original case that only a nuclear engine could move heavy payloads around efficiently. Once again, a working idea was parked not because it could not fly, but because the mission case did not hold up against the budget.

Why the idea keeps coming back

Nuclear propulsion has now been revived and set aside more than once, and there are real reasons it keeps returning and real reasons it keeps stalling. The pull is simple physics: for sending heavy things, and eventually people, on long trips across the solar system, a big jump in efficiency is genuinely valuable.

The obstacles are just as persistent. The engine demands materials that can survive blisteringly hot hydrogen, and hydrogen itself is difficult to store for the months of a Mars voyage without it slowly boiling away. Launching and operating a reactor raises safety and political questions, even though the reactor is flown cold and stays only mildly radioactive until it is switched on in space. And all of it costs money that competes with cheaper alternatives.

What to watch

For now, the nearest-term flight demonstration is off the table, and the question is whether nuclear propulsion returns yet again when Mars ambitions firm up, or whether cheaper chemical launch and other approaches make the case for it later rather than sooner. The thing to watch is whether any agency or company recommits to actually flying a nuclear engine, closing the gap the 1960s left open. The technology that was proven on the ground half a century ago is still, remarkably, waiting for its first trip to space.