NASA, DARPA to launch nuclear rocket to orbit by early 2026

illustration of a nuclear-powered spacecraft in orbit, with earth in the background
Artist's concept of the DRACO (Demonstration for Rocket to Agile Cislunar Operations) spacecraft, which will demonstrate a nuclear thermal rocket engine. Nuclear thermal propulsion technology could be used for future NASA crewed missions to Mars. (Image credit: DARPA)

NASA and the U.S. military plan to launch a nuclear-powered spacecraft to Earth orbit in late 2025 or early 2026.

The project, known as DRACO ("Demonstration Rocket for Agile Cislunar Operations"), aims to give an in-space test to nuclear thermal propulsion (NTP), potentially revolutionary tech that could help humanity set up shop on Mars and other distant worlds.

The DRACO spacecraft will be developed and built by Lockheed Martin, project team members announced today (July 26).

"We're going to put this together, we're going to fly this demonstration, gather a bunch of great data and really, we believe, usher in a new age for the United States [and] for humankind, to support our space exploration mission," Kirk Shireman, vice president of Lockheed Martin Lunar Exploration Campaigns, said during a press conference today.

Related: US military wants nuclear rocket ideas for missions near the moon

DRACO isn't new. The U.S. Defense Advanced Research Projects Agency (DARPA) started the program in 2021, and NASA came aboard in early 2023.

NASA's involvement shouldn't be surprising; the agency's interest in NTP tech goes way back. For example, NASA aimed to launch a crewed Mars mission aboard a nuclear-powered spacecraft by 1979, via a program called NERVA ("Nuclear Engine for Rocket Vehicle Application"). This didn't happen, of course; NERVA was canceled in 1972.

NASA is still shooting for the Red Planet, aiming to get astronauts there by the late 2030s or early 2040s. And it still views nuclear thermal propulsion as a key breakthrough that could make this goal more attainable, by slashing the travel time to and from the Red Planet.

Nuclear thermal rockets carry small fission reactors, which release incredible amounts of heat as they split atoms. This heat is then applied to a propellant gas, which expands and is funneled into space through a nozzle to create thrust.

This process is distinct from the one employed by radioisotope thermoelectric generators (RTGs), nuclear tech that has been flying aboard probes since the early days of the space age. RTGs don't provide propulsion; they harness the heat of radioactive decay to generate electricity, which then powers instruments, motors and other spacecraft gear.

In previous DRACO updates, DARPA and NASA have said they aim to launch the program's first in-space demonstration by 2027. But that timetable may have moved up; Shireman said in today's briefing that the target launch window at the moment is late 2025 or early 2026.

We got other details today as well. For example, Lockheed is partnering with the Virginia-based company BWX Technologies, which will develop the DRACO spacecraft's nuclear reactor and produce its HALEU ("high-assay low-enriched uranium") fuel.

The spacecraft will head to a relatively high orbit around Earth — likely somewhere between 435 and 1,240 miles (700 to 2,000 kilometers), team members said during today's briefing. From such altitudes, it will take at least 300 years for the DRACO demonstrator to fall back to Earth via atmospheric drag — long enough to ensure that all of its nuclear fuel is spent when it comes down.

The mission team will take pains to ensure safety on the way up as well: The DRACO vehicle's nuclear engine will only be activated once it reaches orbit. During launch, the engine will be equipped with a "poison wire," a piece of metal that absorbs neutrons, preventing them from initiating a chain reaction. The poison wire acts much as control rods do in nuclear power reactors here on Earth, team members said. 

DRACO is expected to operate in orbit for a few months. There are no scientific instruments going up; "operation" entails the use of its NTP engine, demonstrating that it can work for long stretches in the space environment.

Use of that engine, however, will also require keeping DRACO's hydrogen — the spacecraft will launch with about 4,400 pounds (2,000 kilograms) of the stuff — super-cold, which is no mean feat.

"Our life-limiting factor is how long we can keep the hydrogen cryogenic," Tabitha Dodson, DRACO program manager at DARPA, said during today's briefing. "This is just as much a demonstration of on-orbit storage of cryogenic liquid hydrogen as it is a demo of the nuclear thermal rocket engine."

Dotson added that, while the spacecraft's specifications are still being worked out, it will basically consist of the NTP engine system and a big tank to hold the hydrogen. (She likened it to "a flying test stand.") And the vehicle won't require a heavy-lift rocket; it will be small enough to fit inside the fairing of a "standard" launcher like SpaceX's Falcon 9.

We also got information about the project's cost. The contracts awarded to Lockheed and BWX Technologies for their DRACO work have a total value of $499 million, provided all milestones are reached, Dotson said. Half of the money will come from DARPA and half from NASA, she added.

Join our Space Forums to keep talking space on the latest missions, night sky and more! And if you have a news tip, correction or comment, let us know at:

Mike Wall
Senior Space Writer

Michael Wall is a Senior Space Writer with and joined the team in 2010. He primarily covers exoplanets, spaceflight and military space, but has been known to dabble in the space art beat. His book about the search for alien life, "Out There," was published on Nov. 13, 2018. Before becoming a science writer, Michael worked as a herpetologist and wildlife biologist. He has a Ph.D. in evolutionary biology from the University of Sydney, Australia, a bachelor's degree from the University of Arizona, and a graduate certificate in science writing from the University of California, Santa Cruz. To find out what his latest project is, you can follow Michael on Twitter.

  • newtons_laws
    A development which is long overdue.
  • Elofson
    Yes, it's about time. People's irrational fear of nuclear energy has stopped development for too long.
  • Unclear Engineer
    I wonder how much advancement has occurred since the early 1970s, when the NERVA fission rocket motor was fired-up on a test stand on Earth. We have learned a lot since then, so I am wondering if Lockheed Martin is just dusting off the NERVA rocket motor, or redesigning the whole thing.

    As for the orbit and the statement that it will not reenter for a period "long enough to ensure that all of its nuclear fuel is spent when it comes down," that is not the real issue. Unfissioned uranium getting vaporized in the atmosphere is not as big a problem as the products of the uranium that did fission getting vaporized in the atmosphere. So, what they need to talk about is the amount of radioactivity in the fission products, not the unfissioned fuel. See , which says " the total radioactivity of a mixture of pure fission products decreases rapidly for the first several hundred years (controlled by the short-lived products) before stabilizing at a low level that changes little for hundreds of thousands of years (controlled by the seven long-lived products)."

    So, the real issue is how much of what nuclear fission product isotopes are going to be released into the atmosphere. There are a lot of different elements in fission waste, some of which are highly bioactive. And, they are not "all gone" after a half-life, some is left for a long time - half of what was left after the previous half-life will still be around after the next half-life.

    Or, worse, what if this rocket motor holds together during reentry, and we end up with all of that "nuclear waste" landing in a big lump somewhere?

    I would like to see an engineered waste control plan for this thing before we launch it.