The Space Test Program: the Pentagon's thirty-million-d...

Just after 2 a.m. on June 25, 2019, a Falcon Heavy lifted off from Kennedy Space Center with two dozen satellites stacked inside its nose cone. Over the next three and a half hours the rocket's upper stage fired four separate times, dropping payloads into three different orbits in a slow orbital ballet that SpaceX called one of the most challenging launches it had ever attempted. The satellites belonged to NASA, to universities, to a nonprofit, and to the Department of Defense. The mission that stitched all of them onto one rocket was called STP-2, and it was run by a Pentagon office most Americans have never heard of. That office is the Space Test Program, and for all the spectacle it puts in orbit, it runs on roughly thirty million dollars a year.

That gap between what the program does and what it costs is the whole story here. Most of the entries in a public-money series are about a big number that bought too little. The Space Test Program is the rarer thing, a small number that has quietly bought a lot, for almost sixty years, without generating the kind of scandal that draws auditors. It is worth understanding precisely because it is the counterexample, the line item that works.

The program that buys rides, not rockets

The Space Test Program is the Department of Defense's primary provider of spaceflight for the military space science and technology community. In plain terms, it is the office that gets experimental government payloads to orbit when those payloads have no ride of their own. A small Air Force lab, a university team on a defense grant, or a service research project can build a genuinely useful sensor or spacecraft and still have no way to reach space, because a dedicated launch costs far more than the experiment itself. STP exists to close that gap. It aggregates these orphan payloads, finds them room on a rocket or on the International Space Station, and flies them, mostly as secondary or rideshare payloads riding along with something bigger.

The lineage runs back to 1965. That year the Director of Defense Research and Engineering authorized a tri-service Space Experiment Support Program and named the Air Force as the executive agent to run it. The effort was later renamed the Space Test Program, and today it is executed by the U.S. Space Force through Space Systems Command. Since those first flights it has built up a long record. A Department of Defense STP overview briefed in 2025 tallies 58 years of experience, 311 missions, and 665 experiments and counting. Those are cumulative lifetime totals, not annual output, and they are the right way to picture the program: a steady trickle of experiments reaching orbit, decade after decade, most of them too small to have made it any other way.

The model matters because it is fundamentally a sharing model. STP does not usually buy its own rocket. It buys space on someone else's, or a slot on the ISS, and packs as many vetted experiments into that space as it can. When it does contract a dedicated launch, as with STP-2, it fills the vehicle with payloads from across the government and beyond. The program's value comes from aggregation, from turning a set of experiments that could never each afford a launch into a single manifest that can.

The review board and the single manager

The gatekeeper in this system is the Space Experiments Review Board, or SERB. Candidate experiments from across the defense research world compete for a spot on it, and the board ranks them, which is how a finite number of ride opportunities gets matched to a much larger pool of hopeful payloads. The SERB first met in January 1966, and it has been the front door to a DoD ride ever since. If your experiment makes the list, STP goes looking for a way to fly it. If it does not, it stays on the ground.

STP's authority is best described as executive-agent-like, and it is worth being precise about what actually designates it, because the paperwork is not what you might expect. The Air Force was named executive agent for military space experimentation back in that 1965 authorization, and the Department of Defense Executive Agent registry maintained by the Office of the Secretary of Defense is the place of record for the designation itself. What does not exist is a single, tidy, numbered DoD Directive titled after the Space Test Program. The program's governance instead rests on a stack of older authorities: Executive Orders 10521 and 13185, DoD Instruction 3210.1 on the administration and support of basic research, a Deputy Secretary of Defense policy memo on Space Test Program management and funding issued in July 2002, and Air Force Instruction 10-1202 that implements all of it. Through that stack, STP functions as the DoD single manager for space experimentation and for all Department of Defense payloads flown on the International Space Station. So when this article calls STP an executive agent, that is shorthand for a real but distributed authority, not a citation to a directive number that does not exist.

The money: thirty million dollars a year

Here is the load-bearing figure, and it is small. The Space Force research, development, test, and evaluation budget carries the Space Test Program as Program Element 1206864SF, R-1 line number 49, inside Budget Activity 6, which is RDT&E management support. In the President's Budget for fiscal year 2026, that line shows $29,121,000 actually spent in FY2024, $30,279,000 enacted for FY2025, and $28,787,000 requested for FY2026. These are the numbers that carry the program's entire reputation as cheap and high-value, and they check out against the primary budget justification.

A few things are worth stating plainly about that figure so it is not misread. This is an annual appropriation line, not a contract ceiling, not a lifetime obligation, and not a multi-year projection. The single project inside the program element, a free-flyer spacecraft line, carries the identical amounts, so the total is not an aggregation of several hidden sub-programs. The small year-over-year dip from FY2025 to FY2026 is not a policy retreat; the justification attributes it to a modest reduction in advisory and assistance services and a transfer of about one million dollars out to the small-business innovation research set-aside. One bookkeeping note for anyone chasing the primary source: older budget books listed this program element as 1206864F under the Air Force, and the current books use 1206864SF under the Space Force, the same program under a new owner. The budget document itself sits with the Department of the Air Force, which as of the 2025 reorganization operates under the department that was renamed the Department of War, though the Air Force and Space Force channels are where this particular justification lives.

Now hold that thirty-million-dollar annual line firmly apart from the other number people cite for STP. The flagship STP-2 rideshare in June 2019 deployed 24 satellites totaling roughly 3,700 kilograms on a Falcon Heavy, across four upper-stage burns over about three and a half hours. Press accounts valued that mission at close to $750 million when you count all the payloads from all the participating agencies, while the SpaceX launch itself had been contracted at around $165 million back in 2012. That $750 million is a one-off, all-in, multi-agency mission value spread across NASA, the Pentagon, universities, and others. It is not STP's yearly appropriation, and it is not even mostly STP's money. Conflating the two would turn a genuinely lean program into a phantom big spender. The honest annual figure is the roughly thirty million in the RDT&E line, and everything else is the cost of the rockets and payloads that STP helps arrange to share.

The rideshare tax: waiting a decade for a ride

The honest critique of STP is not fraud and it is not bloat. It is schedule risk, and it is baked directly into the rideshare model. An experiment is only as timely as the ride it can catch. When you refuse to buy your own rocket and instead wait for a shared slot, you accept that your launch date is hostage to someone else's vehicle, someone else's readiness, and someone else's delays.

The clearest illustration is the DSX satellite, the Demonstration and Science Experiments spacecraft built by the Air Force Research Laboratory to study radiation and the space environment. DSX was originally planned to fly around 2008. Instead it waited roughly a decade for a launcher. It was reassigned from an early ride on a weather satellite mission over to STP-2, and then it sat through the long development of the Falcon Heavy before finally reaching orbit on that June 2019 flight. A good satellite that arrives in orbit a decade late has already burned much of its technological relevance and a real chunk of its design life before it ever switches on. The instruments were state of the art when they were conceived and merely adequate by the time they flew. That is the tax the rideshare model quietly charges, paid not in dollars but in years.

It is only fair to flag the other side of the ledger even inside the critique. No Government Accountability Office report, no inspector general finding, and no Congressional Budget Office study singles out the Space Test Program as wasteful. The watchdog record on this program is thin rather than damning, which is itself a kind of evidence. A program that had been quietly squandering money for sixty years usually leaves an audit trail. This one has not. The problem with STP is timeliness, not integrity.

What thirty million dollars buys

Set the schedule complaint against what the program has actually delivered, and the case for it is strong. For a line item that never cracks thirty-one million dollars a year, STP has punched far above its weight.

The most durable example is a piece of hardware you have probably never heard named. On STP-1 in 2007, the program flew the EELV Secondary Payload Adapter, the ESPA ring, a circular structure that bolts between a primary satellite and its rocket and provides mounting points for a cluster of smaller payloads. That ring worked, and it became the industry standard. The ESPA adapter is now used across commercial and government launches as the default way to fly rideshare payloads. A thirty-million-dollar experimental program invented a component that the entire modern smallsat launch economy now runs on. That is an unusually large return on a modest bet.

Go back further and the payoff is even larger. Early Space Test Program flights carried the experiments that led to the GPS constellation, the satellite navigation system that now underwrites everything from precision munitions to the timestamp on a credit-card transaction. STP did not build GPS, but it flew the early pieces that proved the concept, at a time when nobody would have funded a dedicated launch for them. That is exactly the role the program is designed to play: to give a promising idea a first ride before anyone is willing to bet a rocket on it.

Stack it up and the value proposition is clear. For under thirty million dollars a year, STP gives cash-strapped defense labs, service research projects, and university teams a credible path to orbit, aggregates their experiments onto shared launches, and has done so for hundreds of payloads across nearly six decades. By the cost standards of defense space, where a single dedicated launch can run into the hundreds of millions, that is genuinely cheap and genuinely enabling. It is the difference between an experiment reaching space and an experiment staying on a shelf.

Reading the ledger

Two true things sit side by side here, and neither cancels the other. The Space Test Program is slow, and the Space Test Program is a bargain. The rideshare model that makes it cheap is the same model that left DSX waiting a decade for a launcher and arriving with much of its relevance spent. You cannot have the thirty-million-dollar price tag without accepting that your payloads fly when a ride appears, not when the science is ready.

But the ledger does not read as waste. It reads as a deliberate trade, made openly and priced honestly in the budget books at $30,279,000 enacted for FY2025 and $28,787,000 requested for FY2026, with the mission scale kept where it belongs, cumulative and separate from the annual line. In exchange for tolerating delay, the government gets an ESPA ring the whole launch industry adopted, the early flights behind GPS, and 665 experiments reaching orbit over 58 years that would mostly never have flown at all. Most public-money stories end with a reader asking where the money went. This one ends with a quieter and more uncomfortable question for the taxpayer who wants efficiency: what does it say that the program hardest to criticize is also one of the smallest, and one almost nobody has heard of.

Fact-check notes and sources

Designation and authorities: The Air Force was named executive agent for military space experimentation in 1965 by the Director of Defense Research and Engineering, a role now executed by the Space Force through Space Systems Command. There is no dedicated numbered DoD Directive for the Space Test Program; the governing authorities are Executive Orders 10521 and 13185, DoD Instruction 3210.1, a July 2002 Deputy Secretary of Defense policy memo on STP management and funding, and Air Force Instruction 10-1202, with the executive-agent designation itself recorded in the OSD registry. DoD Executive Agent registry, Space Test Program overview (Wikipedia).
Annual budget (load-bearing): Program Element 1206864SF, RDT&E Space Force, R-1 line 49, shows FY2024 actual $29,121,000, FY2025 enacted $30,279,000, and FY2026 requested $28,787,000. This is an annual appropriation line in RDT&E management support, not a contract ceiling, an obligation, or a life-cycle projection; the older Air Force designator was 1206864F. Verified against the FY2026 Air Force/Space Force RDT&E budget justification (Exhibit R-2, June 2025) and independently reported by an aggregator. HigherGov budget record for the Space Test Program.
Mission scale (cumulative, not annual): A Department of Defense Space Test Program overview briefed in 2025 reports 58 years of experience, 311 missions, and 665 experiments, with the program established in 1965 and the Space Experiments Review Board first meeting in January 1966. NASA-hosted DoD STP overview, 2025 (PDF).
STP-2 mission value (one-off, approximate, not the annual line): The June 2019 STP-2 rideshare deployed 24 satellites totaling about 3,700 kg on a Falcon Heavy across four upper-stage burns over roughly three and a half hours. Press accounts placed the all-in, multi-agency mission value near $750 million counting all payloads, with the launch itself contracted around $165 million in 2012. These are approximate one-off figures spread across multiple agencies and are kept separate from STP's annual appropriation. SpaceNews, Falcon Heavy launches STP-2.
Schedule risk and the payoff: The AFRL-built DSX satellite was planned to launch around 2008 but did not fly until STP-2 in June 2019, after being reassigned launchers and delayed by Falcon Heavy development. On the value side, STP flew the first ESPA rideshare ring on STP-1 in 2007 (now the industry-standard adapter) and its early flights carried experiments that led to the GPS constellation. No GAO, IG, or CBO report singles out STP as wasteful. Demonstration and Science Experiments (Wikipedia), Space Test Program (Wikipedia).

This post is informational and journalistic, drawn from public records, and is not legal, financial, or policy advice; all dollar figures are attributed to their stated fiscal year.

The Space Test Program: the Pentagon's thirty-million-dollar ride to orbit