NASA Rolls Out Artemis III Moon Rocket Core Stage Ahead of Next Lunar Mission

On April 20, 2026, workers at NASA's Michoud Assembly Facility in New Orleans guided the upper four-fifths of a 212-foot rocket core stage onto the agency's Pegasus barge for a 900-mile voyage across the Gulf of Mexico to Kennedy Space Center [1]. The hardware — comprising the liquid hydrogen tank, liquid oxygen tank, intertank, and forward skirt of the Space Launch System — will eventually power Artemis III, the mission that was supposed to return American astronauts to the lunar surface for the first time since 1972 [2].

Instead, Artemis III will orbit Earth.

Two months before the rollout, on February 27, 2026, NASA Administrator Jared Isaacman confirmed that Artemis III had been fundamentally redesigned. Rather than landing a crew at the Moon's south pole, the mission will now conduct rendezvous and docking tests in low Earth orbit with one or both commercially developed lunar landers — SpaceX's Starship Human Landing System and Blue Origin's Blue Moon [3]. The actual crewed lunar landing has been deferred to Artemis IV, tentatively scheduled for 2028 [4]. The SLS core stage rolling out of Michoud will carry astronauts no closer to the Moon than the International Space Station's altitude.

A Mission That Keeps Changing

The original Artemis III plan, announced in 2019, was ambitious: launch two astronauts to lunar orbit aboard SLS and Orion, transfer them to a commercial lander, and put boots on the Moon near the south pole for roughly a week, with a total mission duration of about 30 days [5]. The target date was late 2024.

That date has slipped six times since.

Source: NASA / SpaceNews / Wikipedia

Data as of Apr 21, 2026CSV

In November 2021, NASA Administrator Bill Nelson acknowledged the timeline had slipped to 2025 [6]. By June 2023, officials said "probably" no earlier than 2026 [7]. In January 2024, NASA officially moved the date to September 2026 [8]. In December 2024, it slipped again to mid-2027 [9]. Then in January 2026, Isaacman's NASA pushed the crewed landing to 2028 under the new Artemis IV designation, while accelerating a stripped-down Artemis III to mid-2027 as an orbital test flight [3].

The delays trace to multiple overlapping technical failures. The Orion heat shield suffered unexpected "char loss" during Artemis I's return from the Moon in December 2022, when gases generated inside the ablative Avcoat material could not vent properly, causing cracking and material loss [10]. NASA spent over a year diagnosing the root cause: the heat shield was not porous enough, and the "skip entry" reentry trajectory — designed to improve landing precision — exacerbated pressure buildup within the shield's decomposition layer [10]. Rather than redesigning the shield, NASA modified the reentry trajectory for Artemis II, eliminating the skip maneuver for a steeper, shorter-range approach [11]. That fix appears to have worked — Artemis II splashed down successfully on April 1, 2026, with commander Reid Wiseman reporting the heat shield "looked wonderful" [12].

Spacesuit development compounded the timeline problems. Axiom Space's Extravehicular Mobility Unit (AxEMU), the suit astronauts will need for lunar surface operations, is at least a year and a half behind schedule, with demonstration readiness not expected until late 2027 [6]. Since Artemis III no longer involves a lunar landing, the AxEMU will instead be tested in the more forgiving environment of low Earth orbit [3].

The Recurring Hydrogen Problem

The SLS core stage traces its lineage directly to the Space Shuttle's External Tank, and it has inherited a persistent engineering challenge: liquid hydrogen leaks [13]. During Artemis I's launch campaign in 2022, hydrogen leaks at a quick-disconnect fitting forced multiple scrubs before engineers developed a workaround involving thermal cycling of the seals [14]. Similar leaks appeared during Artemis II's wet dress rehearsal in February 2026, forcing engineers to halt hydrogen loading and apply the same troubleshooting procedures developed after the Artemis I experience [15].

The Artemis III core stage now being shipped to Kennedy Space Center is structurally similar to its predecessors. Boeing, the lead contractor for core stage manufacturing, and L3Harris Technologies, responsible for the RS-25 engines, have applied lessons from the first two missions [1]. The four RS-25 engines for Artemis III are scheduled to ship from NASA's Stennis Space Center no later than July 2026 for integration into the engine section already at Kennedy [1]. But no independent public verification has confirmed that the hydrogen leak issues have been structurally resolved rather than operationally managed — the fixes to date have centered on ground procedures and thermal techniques rather than hardware redesign [14][15].

$31.6 Billion and Counting

The SLS program's price tag has grown relentlessly. Initial congressional authorization in 2010 envisioned a vehicle that would cost roughly $10 billion to develop. By 2022, cumulative spending had reached $23 billion [16]. As of 2025, the figure stands at $31.6 billion [17], and the program has produced exactly two flights: one uncrewed test and one crewed lunar flyby.

Source: NASA OIG / Wikipedia

Data as of Apr 21, 2026CSV

Each Artemis launch — including SLS, Orion, and Exploration Ground Systems — costs approximately $4.1 billion, according to NASA's own figures confirmed by acting administrator Sean Duffy [18]. The broader Artemis campaign, including lander contracts, Gateway development, and ground infrastructure, is projected to cost $93 billion through 2025 [19].

For context, the Apollo program's per-mission cost averaged $2.73 billion in 2025 dollars [20]. Artemis missions cost roughly 50% more per launch than Apollo did — and Apollo actually landed on the Moon.

Source: Planetary Society / NASA OIG / SpaceX estimates

Data as of Apr 21, 2026CSV

At the other end of the cost spectrum, SpaceX's Starship, which is designed for full reusability, has a projected operational cost that some estimates place as low as $2 million per launch once the vehicle achieves routine reuse [21]. NASA's fixed-price contract with SpaceX for the Starship HLS lunar lander, including development and two operational flights, totals $2.89 billion — less than the cost of a single SLS launch [22]. A supplemental Option B contract for a second-generation Starship HLS design added $1.15 billion [22]. The total SpaceX lander contract, covering design, manufacture, an uncrewed demo, and a crewed landing, costs less than one SLS flight to low Earth orbit.

NASA's Office of Inspector General has been blunt about the cost trajectory. A 2023 report concluded that NASA's projection of a 50% reduction in SLS launch costs through a transition to a services contract was "highly unrealistic," and that per-vehicle costs would likely remain above $2 billion for the foreseeable future [23]. The OIG recommended NASA "continue to monitor the commercial development of heavy-lift space flight systems and begin discussions of whether it makes financial and strategic sense to consider these options" [23].

The Political Architecture of SLS

The SLS program was not born from a blank-page engineering analysis. It was created through the 2010 NASA Authorization Act, explicitly designed to reuse Space Shuttle infrastructure and preserve contractor relationships across key congressional districts in Alabama, Louisiana, Texas, and California [24]. Critics have long labeled it the "Senate Launch System" for this reason.

The program supports approximately 20,000 direct jobs across multiple states [24]. Boeing's SLS program is managed from Huntsville, Alabama, with additional employees at Michoud in New Orleans and other facilities nationwide [25]. Rep. Dale Strong (R-Ala.), whose district includes NASA's Marshall Space Flight Center — the lead center for SLS — serves as vice-chair of the commerce, justice, and science subcommittee of the House Appropriations Committee [24]. This structural alignment between congressional appropriators and SLS employment centers has proven durable across multiple administrations.

In early 2025, Boeing warned SLS employees of potential layoffs, with approximately 200 positions at risk as the company anticipated possible program cuts [25]. But the program has survived every previous cancellation threat, in part because of the political cost of eliminating specialized aerospace jobs in electorally important states.

The Trump Administration's FY2027 budget request proposed replacing SLS with commercial transportation services after Artemis V [26]. Congress subsequently funded Artemis IV and V through the One Big Beautiful Bill Act of July 2025, but directed NASA to study commercial alternatives to the Exploration Upper Stage [26]. These are incremental steps that acknowledge the cost problem without resolving it.

What the Mission Will — and Won't — Demonstrate

Under the revised plan, Artemis III will test several capabilities essential to future lunar landing missions: rendezvous and docking between Orion and at least one commercial lander (potentially both Starship HLS and Blue Moon), propulsion and life support systems aboard the landers, communication systems, and the AxEMU spacesuit in a spacewalk from the lander [3][4].

These are genuine engineering milestones. Orbital rendezvous and crew transfer between vehicles is a critical operation for any lunar architecture that uses separate launch and landing vehicles, and testing it with human crews reduces risk for subsequent missions. The dual-lander test — if both SpaceX and Blue Origin vehicles are ready — would provide NASA with operational data on two independent landing systems, creating redundancy in the program's most critical single point of failure [4].

However, none of these tests require SLS. The Orion spacecraft could, in principle, be launched on a commercial heavy-lift vehicle. The lander tests require the landers to reach orbit, which they must do on their own launch vehicles regardless. The spacesuit test could be conducted on any crewed orbital platform. The unique capability SLS provides — throwing Orion on a translunar injection trajectory — is not being used on Artemis III at all [3].

The science payloads originally selected for Artemis III's lunar surface operations — including the Lunar Environment Monitoring Station (LEMS), a seismometer to characterize the Moon's crust and mantle; Lunar Effects on Agricultural Flora (LEAF), an experiment on space crop cultivation; and the Lunar Dielectric Analyzer (LDA), an internationally contributed instrument to study lunar regolith — will need to wait for Artemis IV or later missions [27]. None of these instruments can be deployed from Earth orbit.

The Case For and Against Continuing SLS

The strongest argument for continuing SLS comes not from NASA officials but from the structural reality of the Artemis architecture. Orion was designed for SLS. The launch tower, mobile launcher, and ground systems at Kennedy Space Center were built for SLS. Switching to a commercial launch vehicle would require redesigning the spacecraft-to-rocket interface, recertifying the vehicle for human flight on a new launcher, and potentially rebuilding ground infrastructure — costs and delays that could rival continuing SLS for the remaining planned missions [26].

Proponents also argue that government-owned launch infrastructure provides strategic independence from commercial providers. Relying exclusively on SpaceX for both the lander and the crew launch vehicle would create a single-vendor dependency that introduces risks from corporate financial instability, labor disputes, or strategic decisions that conflict with government priorities [26].

Critics counter that these arguments amount to a sunk-cost fallacy. Aerospace analyst Casey Dreier of the Planetary Society has documented that Artemis spends roughly $6 billion per year in inflation-adjusted dollars, compared to Apollo's peak of $42 billion per year — but produces results at a fraction of Apollo's pace [20]. The OIG has repeatedly flagged Boeing's "poor performance" as a driver of cost growth, citing a 2018 audit that attributed most delays to "management, technical, and infrastructure issues" at the prime contractor [16]. When the OIG recommended financial penalties for Boeing's noncompliance with quality standards, NASA responded that such penalties were "outside the bounds of the contract" [16] — an answer that raises questions about the contract structure itself.

The International Dimension

Sixty-two nations have now signed the Artemis Accords, with Latvia joining on April 20, 2026 [28]. But the accords are primarily statements of principle regarding transparency, interoperability, and space resource use — they do not constitute binding financial commitments to specific mission timelines [28].

The hardware commitments are more concrete and more vulnerable to schedule slips. The European Space Agency has committed €787.5 million in Gateway hardware contracts, including the I-Hab module built by Thales Alenia Space [29]. Canada awarded a $999.8 million contract to MDA for the Canadarm3 robotic system [30]. Japan's JAXA is providing environmental control, life support, batteries, and thermal control systems for the I-Hab [31].

However, on March 24, 2026, NASA announced it was pausing Gateway development to focus on building a lunar surface base instead [29]. This decision throws the timeline for international hardware contributions into uncertainty and risks straining relationships with partners who have already committed substantial public funds.

The geopolitical stakes extend beyond partner management. China's 2030 target for a crewed lunar landing, pursued through its partnership with Russia on the International Lunar Research Station (which now has 13 partner countries, none of whom have signed the Artemis Accords), creates direct competitive pressure [32][33]. Every year of Artemis delay narrows the gap — or eliminates it entirely. If China lands astronauts on the Moon before the United States returns, it would represent a symbolic shift with real diplomatic consequences for nations deciding between the Artemis and ILRS frameworks.

What Happens Next

The Pegasus barge carrying the Artemis III core stage section will take six to eight days to reach Kennedy Space Center [1]. Once there, technicians will mate the newly arrived structure with the engine section and boat-tail already in the Vehicle Assembly Building [2]. The four RS-25 engines will follow from Stennis by July [1]. Full vehicle integration will continue through late 2026 and into 2027, with a targeted launch in mid-2027 [3].

Whether the mission flies on schedule depends on factors largely outside the core stage's control: Starship HLS and Blue Moon must be ready for orbital operations, Axiom must deliver flight-ready spacesuits, and the ground systems at Kennedy must be refurbished after Artemis II [4].

The core stage rolling out of Michoud represents real engineering achievement — a 212-foot cryogenic propulsion system that will generate over 8 million pounds of thrust at liftoff. But it also represents a program that has spent $31.6 billion to fly twice, is about to fly a mission stripped of its original purpose, and faces an uncertain future beyond the next two or three flights. The rocket works. The question is whether it is the right rocket, at the right price, for what the United States is trying to accomplish on the Moon.