NASA Artemis II Capsule Returns to Launch Site; Artemis III Hardware Begins Arriving at Kennedy Space Center

On April 10, 2026, the Orion spacecraft splashed down in the Pacific Ocean southwest of San Diego, landing just 2.9 miles from its targeted point — the first crewed NASA capsule recovered by the U.S. Navy since the Apollo-Soyuz Test Project in 1975 [1]. Astronauts Reid Wiseman, Victor Glover, Christina Koch, and Jeremy Hansen had traveled 694,481 miles around the Moon and back, proving that the hardware could keep humans alive in deep space [2].

Seventeen days later, on April 27, a barge named Pegasus docked at Kennedy Space Center carrying the 212-foot core stage of the SLS rocket designated for Artemis III [3]. The symbolism was tidy: one mission's hardware coming home for study, the next mission's hardware arriving for assembly. But between those two events lies a gap filled with budget fights, technical unknowns, and a fundamental redesign of what Artemis III will actually do.

The Mission That Lost Its Landing

The original plan called for Artemis III to put astronauts on the lunar surface — the first boots on the Moon since Apollo 17 in 1972. That is no longer happening.

On February 27, 2026, NASA Administrator Jared Isaacman confirmed that Artemis III has been restructured into an orbital demonstration mission [4]. Instead of landing on the Moon, astronauts will rendezvous and dock with one or both commercial lunar landers — SpaceX's Starship Human Landing System and Blue Origin's Blue Moon — in Earth orbit. They will test propulsion, life support, communications, and the new Axiom Extravehicular Mobility Unit (AxEMU) spacesuits [4]. The actual crewed lunar landing has been pushed to Artemis IV, tentatively set for 2028 [5].

Mission planners are still deciding between a low Earth orbit (LEO) and high Earth orbit (HEO) profile for Artemis III. A LEO mission could preserve an Interim Cryogenic Propulsion Stage for Artemis IV, while a HEO mission would more closely simulate thermal and operational conditions near the Moon [4]. As of late April 2026, NASA had not finalized the orbital profile or confirmed whether the mission would dock with one or both HLS vehicles [4].

The timeline gap tells its own story. From Artemis II's splashdown on April 10, 2026, to Artemis III's current target of late 2027, roughly 18 months will have elapsed [6]. From there, the first lunar landing attempt on Artemis IV is projected for 2028 — meaning at least two years between a successful crewed lunar flyby and an attempted surface mission. During the Apollo era, NASA flew Apollo 10 (lunar orbit) in May 1969 and Apollo 11 (first landing) just two months later [7]. The comparison is imperfect — Apollo's missions used a single, proven lander architecture — but it underscores how much more cautious (or constrained) the current program is.

Following the Money: A $66 Billion Program and Counting

The Artemis program's cumulative spending through 2025 reached approximately $66 billion, broken down across its major components [8][9].

Source: The Planetary Society / NASA OIG

Data as of Apr 1, 2026CSV

The SLS rocket alone consumed $27.2 billion through 2025, with cost overruns of 42.5% above original projections [8]. Orion exceeded its estimates by 37.4%, and Exploration Ground Systems ran 40% over budget, which NASA's Office of Inspector General attributed to "poorly defined requirements, poor contractor performance, and increased material cost" [8].

Annual spending continues at a rapid clip. In fiscal year 2025, NASA allocated $2.6 billion for SLS, $1.3 billion for Orion, and $900 million for ground systems — a combined $4.8 billion per year just to maintain the program's core elements [8].

Each SLS launch costs approximately $4 billion — four times the original estimate of roughly $1 billion per flight [9][10]. That figure dwarfs every alternative on the market and even surpasses the inflation-adjusted cost of a Saturn V launch during Apollo.

Source: NASA OIG / Industry Estimates

Data as of Apr 1, 2026CSV

SpaceX's Falcon Heavy launches for roughly $150 million. Starship, if it achieves full reusability, projects a per-launch cost under $100 million [10]. Even the Saturn V, adjusted to 2025 dollars, cost approximately $1.4 billion per flight [11]. The SLS costs nearly three times as much as the rocket that first put humans on the Moon.

For broader context, Apollo's total program cost through its first lunar landing was approximately $290 billion in 2025 dollars, but that reflected annual spending peaks of $42 billion (adjusted) — about seven times Artemis's annual budget of roughly $6 billion [11]. Artemis is attempting to achieve similar results at a fraction of the yearly expenditure, which helps explain the extended timelines.

Starship's Unfinished Homework

The entire Artemis surface architecture depends on SpaceX's Starship HLS functioning as a crewed lunar lander. As of April 2026, several critical milestones remain unmet.

SpaceX has completed 49 HLS-specific milestones tied to subsystem development, covering power generation, communications, guidance, navigation, propulsion, life support, and radiation protection [12]. But the single most consequential demonstration — ship-to-ship cryogenic propellant transfer in orbit — has not yet occurred [13].

The Starship HLS architecture requires launching the lander, then refueling it in orbit through multiple tanker flights before it can transit to the Moon. SpaceX tested internal propellant transfer during Integrated Flight Test 3 in March 2024 [13], but the ship-to-ship transfer demo originally planned for mid-2025 slipped and is now expected sometime in 2026 [13]. Until that test succeeds, NASA cannot certify the system for crewed operations.

SpaceX has also committed to a long-duration orbital flight test of the HLS variant in 2026 [12]. Both SpaceX and Blue Origin have told NASA they can support a late-2027 rendezvous and docking test ahead of a landing attempt in 2028 [4], but neither has demonstrated their lander in the relevant environment.

If Starship misses these milestones, the downstream schedule effects are straightforward: Artemis IV's 2028 landing target slips, and NASA faces the uncomfortable position of having a crew-rated capsule and rocket with nowhere to land.

Who Flies, and What They'll Do on the Surface

NASA has not publicly named an Artemis III crew. The Artemis II crew — Wiseman, Glover, Koch, and Hansen — completed their mission successfully, but crew assignments for subsequent missions depend on the finalized mission profile [2].

For the eventual landing mission (now Artemis IV), the original Artemis III surface plan called for two astronauts to descend to the lunar south pole aboard the Starship HLS while two remained in Orion in lunar orbit. The surface crew was to conduct two to five extravehicular activities during a 6.5-day stay — approximately 156 hours on the surface [14].

By comparison, Apollo 11's crew spent just 21 hours and 36 minutes on the surface with a single 2.5-hour EVA. Apollo 17, the longest surface mission, logged about 75 hours with three EVAs totaling 22 hours [7]. Even the most conservative Artemis surface plan would more than double Apollo 17's total surface time.

The Political Geography of Artemis

The Artemis program touches every state in the union, and that is by design.

NASA's prime contractors — Boeing, Lockheed Martin, Northrop Grumman, SpaceX, Blue Origin, Axiom Space, and others — maintain more than 2,700 suppliers across 47 states [15]. Boeing builds the SLS core stage at NASA's Michoud Assembly Facility in New Orleans. Lockheed Martin assembles Orion in Denver. Northrop Grumman produces the solid rocket boosters in Promontory, Utah. SpaceX develops Starship HLS at its facilities in Hawthorne, California, and Boca Chica, Texas [15].

California alone contributes more than 500 companies and 16,000 direct workers to the Artemis program, with each NASA job supporting an additional 35.7 jobs statewide — a total employment impact the governor's office estimates at 66,208 jobs [16].

This geographic spread is not incidental. The SLS was mandated by the NASA Authorization Act of 2010, which specified use of Space Shuttle-derived components "where possible" — a provision that effectively locked in existing contractors and their workforce in specific congressional districts [10]. Cost-plus contracts, under which contractors are reimbursed for all expenses plus a guaranteed profit margin, created what critics call a structural incentive to "hire as much as possible and work as slowly as possible" [9].

Senator Ted Cruz of Texas, whose state hosts NASA's Johnson Space Center, introduced a legislative directive in 2026 to add nearly $10 billion to a budget reconciliation bill, restoring funding for Artemis IV and V, the International Space Station, and direct allocations to Marshall Space Flight Center in Huntsville, Alabama [17]. The directive followed the Trump administration's fiscal year 2026 budget proposal, which called for terminating SLS after Artemis III, describing it as "grossly expensive" [10].

The tension between the executive branch's cost concerns and Congress's employment interests has defined the program for over a decade. NASA's own Inspector General has repeatedly flagged the SLS as unsustainable at current cost levels, yet the program has survived budget pressure from both Republican and Democratic administrations [8][9].

The Case Against SLS — and the Cost of Walking Away

Critics of the SLS argue that it is an expendable rocket in an era of reusability, consuming roughly $2.5 billion per year in production and operations costs that could fund dozens of commercial launches [10]. The Falcon Heavy can lift 64 metric tons to LEO for $150 million; Starship aims for 100–150 metric tons at under $100 million [10]. SLS Block 1 lifts 95 metric tons to LEO for $4 billion.

Administrator Isaacman has signaled agreement. "It's not the vehicle that you are going to take to and from the moon a couple of times a year," he said of the SLS after Artemis II [18]. He canceled Boeing's contract for the SLS Exploration Upper Stage, paused the Gateway lunar orbiting station, and solicited alternative proposals from SpaceX and Blue Origin [18].

But transitioning away from SLS carries its own costs. Over $50 billion in development spending would become sunk costs [8]. No commercial rocket has yet demonstrated the capability to send Orion-class payloads beyond Earth orbit. And Congress — which writes NASA's budget — has shown little appetite for abandoning the jobs and contracts tied to SLS production.

NASA's independent review boards have walked a middle line: acknowledging SLS's cost problems while noting that no alternative system is currently certified for the deep-space crewed missions Artemis requires [8].

Science at the South Pole

The eventual Artemis landing missions target the lunar south pole, where NASA has identified nine candidate landing regions [19]. The leading sites cluster near Shackleton and de Gerlache craters, chosen for their geological diversity and proximity to permanently shadowed regions (PSRs) — craters that have not seen sunlight in billions of years and are believed to contain significant deposits of water ice [19].

The scientific objectives are threefold. First, collect samples from PSRs to determine the depth, distribution, and composition of lunar water ice — data that could determine whether in-situ resource utilization (extracting water for drinking, oxygen production, and hydrogen fuel) is viable for sustained lunar operations [19]. Second, study ancient lunar rocks and regolith that may preserve a record of the early solar system, including bombardment history and volatile delivery from comets and asteroids [19]. Third, test technologies for extended surface operations that would inform future Mars missions.

Critics question whether these objectives require a crewed mission at this stage. Robotic landers — several of which NASA is already funding through the Commercial Lunar Payload Services (CLPS) program — could collect samples and prospect for water ice at a fraction of the cost. Proponents counter that human geologists can make real-time decisions about sampling locations, adapt to unexpected findings, and cover far more terrain than any current rover [19].

Radiation, Risk, and Transparency

Artemis II provided the first real-world crewed radiation data for a lunar-distance mission since Apollo. The results raised eyebrows.

Each crew member accumulated approximately 12 millisieverts (mSv) of radiation exposure over the 10-day mission — 18% higher than NASA's pre-flight models predicted [20]. For context, the average American absorbs about 6.2 mSv per year from natural and medical sources, and a single chest X-ray delivers about 0.1 mSv [20].

Dr. J.D. Polk, NASA's chief health officer, characterized the variance as "significant from a planning perspective, even though it does not represent an immediate health concern for this crew" [20]. The elevated readings resulted from the crew venturing beyond Earth's magnetosphere during a period of solar maximum, which produced unexpected radiation dynamics from both galactic cosmic rays and solar energetic particles [20].

NASA is now accelerating development of advanced shielding technologies and conducting updated risk assessments before approving future deep-space missions [20]. For the Artemis surface missions, which will expose astronauts to unshielded lunar conditions during EVAs, the radiation question becomes more pressing.

The agency's safety architecture includes several abort options. The SLS Launch Abort System can pull the Orion capsule to safety within seconds of a launch anomaly [21]. The Artemis II mission profile included a free-return trajectory — what NASA calls a "built-in safety net" — allowing the spacecraft to loop around the Moon and return to Earth using gravity alone if the main engines failed [21]. For surface missions, abort scenarios become more complex: an emergency ascent from the lunar surface aboard Starship HLS, orbital rendezvous with Orion, and return to Earth would involve multiple independent systems performing flawlessly in sequence.

NASA has been more forthcoming about radiation data than the Apollo-era norm, publishing exposure figures within weeks of splashdown [20]. But residual unknowns remain, particularly around cumulative effects of galactic cosmic rays during extended surface stays and the biological margins built into career exposure limits.

What Comes Next

The Orion capsule from Artemis II is now at Kennedy Space Center's Multi-Payload Processing Facility, where teams will extract data, remove reusable components, and conduct detailed heat shield inspections [1]. NASA reported that the thermal protection system "performed as expected," with char loss "significantly reduced" compared to the uncrewed Artemis I flight [1]. Over the summer, the heat shield will transfer to Marshall Space Flight Center for x-ray analysis and sample extraction [1].

Meanwhile, in the Vehicle Assembly Building, the Artemis III core stage is being mated with the engine section that arrived in August 2025 [3]. If the late-2027 launch target holds, teams have roughly 18 months to integrate the rocket, stack it on the mobile launcher, and roll it to Launch Complex 39B.

The broader picture is one of a program in transition. The SLS has proven it can fly humans to the Moon and back. Orion's heat shield works. The ground systems at Kennedy can support deep-space launches. But the elements needed for a lunar landing — a certified crewed lander, orbital refueling, surface spacesuits — remain in development. Whether the 2028 landing target survives contact with engineering reality will depend on milestones that SpaceX, Blue Origin, and Axiom Space have yet to demonstrate.

The money, the politics, and the technical risk all converge on a single question: how much longer will taxpayers and Congress sustain a $5-billion-per-year program whose central promise — returning humans to the lunar surface — keeps receding toward the horizon?