We know how to euthanize beloved pets — veterinarians do it every day. And we know how physician-assisted suicide works — it is legal in several states. If drugs can be used to humanely end life in these other contexts, why is it so difficult in the death penalty context? The answer is one of the best-kept secrets of the killing state: lethal injection is not based on science. It is based on the illusion of science, the assumption of science. “What we have here is a masquerade,” one lab scientist says. “Something that pretends to be science and pretends to be medicine but isn’t.” Consider first the birth of lethal injection.
In 1976, the Supreme Court gave states the green light to resume executions after a decade of legal wrangling over the constitutionality of the death penalty, and Oklahoma was eager to get started. The only hitch was how to do it. Oklahoma’s electric chair was dilapidated and in need of repair, but more importantly, it was widely viewed as barbaric and inhumane. The state was looking to try something new. A state legislator approached several physicians about the possibility of death by drugs — a lethal injection. They wanted nothing to do with it, but the state’s medical examiner, Dr. Jay Chapman, was game. “To hell with them,” the legislator remembered Chapman saying. “Let’s do this.”
Chapman had no expertise in drugs or executions. As Chapman himself would later say, he was an “expert in dead bodies but not an expert in getting them that way.” Still, he said he would help and so he did, dictating a drug combination to the legislator during a meeting in the legislator’s office. Chapman first proposed two drugs, then later added a third. Voila. In 1977, the three-drug protocol that states would use for the next 30 years was born.
The idea was triple toxicity — a megadose of three drugs, any one of which was lethal enough to kill. The first drug, sodium thiopental, would kill by barbiturate overdose, slowing respiration until it stopped entirely. The second drug, pancuronium bromide, would kill by paralyzing the diaphragm, preventing it from pumping air into the lungs. And the third drug, potassium chloride, would kill by triggering a cardiac arrest. The effects of the second and third drugs would be excruciatingly painful, so the first drug did double duty by blocking pain as well.
How did Chapman come up with his three-drug combo? “I didn’t do any research,” he later confided in an interview. “I just knew from having been placed under anesthesia myself, what was needed. I wanted to have at least two drugs in doses that would each kill the prisoner, to make sure if one didn’t kill him, the other would.” As to why he added a third drug, Chapman answered, “Why not? … You wanted to make sure the prisoner was dead at the end, so why not add a third drug,” he said, asking: “Why does it matter why I chose it?”
This is how the original three-drug lethal injection protocol came to be: a man working outside his area of expertise and who had done no research just came up with it. “There was no science,” says law professor Deborah Denno, one of the leading experts in the field. “It was basically concocted in an afternoon.” As another lethal injection expert, law professor Ty Alper, put the point, Chapman “gave the matter about as much thought as you might put in developing a protocol for stacking dishes in a dishwasher.” For the careful dish stackers among us, it’s fair to say he gave it less.
But that was good enough for Oklahoma, which adopted the new execution method without subjecting it to a shred of scientific scrutiny. No committee hearings. No expert testimony. No review of clinical, veterinary, or medical literature. The state was embarking upon an entirely new way to kill its prisoners, and did none of the most basic things.
Texas followed Oklahoma’s lead the next day, and then other states did too, carelessly copying a protocol that had been carelessly designed in the first place. “There is scant evidence that ensuing States’ adoption of lethal injection was supported by any additional medical or scientific studies,” a court reviewing the historical record wrote. “Rather, it is this Court’s impression that the various States simply fell in line relying solely on Oklahoma’s protocol.” As Deborah Denno observes, the result was an optical illusion — states touted a “seemingly modern, scientific method of execution” without an iota of science to back it up. Jay Chapman was as surprised as anyone by other states’ adoption of his protocol. “I guess they just blindly followed it,” he later stated, adding, “Not in my wildest flight of fancy would I have ever thought that it would’ve mushroomed into what it did.” “I was young at the time,” he explained. “I had no idea that it would ever amount to anything except for Oklahoma.”
Over time, every death penalty state in the country would adopt Chapman’s three-drug lethal injection protocol — not because they had studied it, but because in the absence of studying it, there was nothing to do but follow the lead of other states. “I didn’t have the knowledge to question the chemicals,” one warden explained, saying that he had “no reason to because other states were doing it.”12 “It wasn’t a medical decision,” an official from another state explained. “It was based on the other states.”
Sociologists have a name for this, a term of art for fads based on a faulty assumption. They call it a “cascade to a mistaken consensus,” and lethal injection is a textbook example. States had come to a consensus in adopting the three-drug protocol, but it was based on the assumption that other states knew what they were doing. They did not.
The fact that the three-drug protocol wasn’t based on science is not to say that science on the drugs didn’t exist. All three drugs were FDA approved, so there were studies and FDA warning labels saying what each drug did. The problem was that none of that science could predict what would happen when the drugs were used in lethal injection. Lethal injection is an “off-label” use of a drug, and although doctors use drugs for off-label purposes all the time, they aren’t trying to kill people, so their off-label use doesn’t come anywhere close to the use of those drugs as poison in lethal injection. Lethal injection uses drugs in amounts that no one has ever prescribed, let alone studied in a research setting. It delivers the entire dose of a drug at once — a practice known as “bolus dosing” — rather than delivering the drug in an IV drip, as is typical for large doses in the clinical setting. And it uses combinations of drugs that are simply unfathomable in the practice of medicine, giving rise to the possibility of “profound physiological derangements” (science-speak for freakishly weird results), as overdoses of different drugs affect the body in different ways.
Who knew what was going to happen when all three of these perversions came together. No one did, and the studies to find out had not even begun. In the biomedical research setting, a baseline showing of scientific support is required for testing on animals, and the three-drug protocol didn’t even meet that threshold. As one lab scientist quipped, “You wouldn’t be able to use this protocol to kill a pig.”
But states weren’t killing pigs. They were killing people, so they forged ahead, undaunted by the unknowns. Yet over time, the executions that followed created data points of their own, and those data points drew scientists. If states would not go to the science, science would come to them.
Granted, the data was thin. In some states, the problem was secrecy. “There is an enormous amount of information from executions (autopsies, toxicology, ECG recordings, EEG recordings, execution logs, and photographs),” one expert explained, “but most of it has been kept secret.” In other states, the problem was poor record-keeping. In still others, it was a state’s decision to stop keeping records altogether. For example, Texas — which conducts more executions per year than any other state — stopped conducting post-execution autopsies altogether in 1989. “We know how they died,” a state spokesperson stated when asked about the reason for the no-autopsy policy.
That said, the raw data that scientists did manage to get was enough to raise serious concerns about the three-drug protocol. State officials were making “scientifically unsupportable” claims about lethal injection, researchers stated, so they decided to look at the data to see what it showed. In 2005 and 2007, researchers published two peer-reviewed studies on lethal injection, the first major studies of their kind.
In the first study, researchers obtained toxicology reports from forty-nine executions in Arizona, Georgia, North Carolina, and South Carolina. (Texas and Virginia, the two states with the most executions in the country at the time, refused to share their data.) Because they had no other way to determine whether prisoners were anesthetized when they were injected with the second and third drugs, researchers measured the postmortem amounts of sodium thiopental (the first drug) in the blood, finding that most prisoners had amounts lower than what was necessary for anesthesia, and some had only trace amounts in their system.
“Extrapolation of ante-mortem depth of anesthesia from post-mortem thiopental concentrations is admittedly problematic,” the researchers conceded. Still, the wide range of sodium thiopental amounts in prisoners’ blood suggested gross disparities during their executions as well. “It is possible that some of these inmates were fully aware during their executions,” the researchers stated, but their conclusion was more modest: “We certainly cannot conclude that these inmates were unconscious and insensate.”
Vigorous debate ensued. “You can’t take these post-mortem drug levels at face value,” one forensic pathologist stated, explaining that the amount of a drug in the blood dissipates after death, just as it does in life, and most autopsies in the study were conducted around twelve hours after death, so the postmortem measurements didn’t say much about the sodium thiopental in a prisoner’s blood during the execution. The study’s authors shot back with point-by-point responses to the criticism, but the damage was done. The so-called “Lancet study,” named for its publication in one of the most prestigious medical journals in the world, would forever be tainted by skepticism.
Had the first study been the only study of the three-drug protocol, one might have said that the science was inconclusive. But a second study was published two years later, and its findings were far less subject to dispute. In the second study, researchers examined execution logs in California. California’s expert had testified that the effects of sodium thiopental were well understood. Within sixty seconds of receiving the overdose, “over 99.999999999999% of the population would be unconscious,” the state’s expert stated, and “virtually all persons [would] stop breathing within a minute.” But when researchers examined the logs from California’s eleven executions by lethal injection, they found that this was not the case. In six of the eleven cases — 54% — the logs showed that the prisoner “continued to breathe for up to nine minutes after thiopental was injected.”
This was alarming not only because it showed that the state’s expert was wrong, but also because it suggested that the prisoners had died torturous deaths. In the absence of a trained professional assessing anesthetic depth, the cessation of breathing provides a rough proxy for adequate anesthesia. Thus, the fact that over half the prisoners continued breathing was an ominous sign that they had not been fully anesthetized prior to injection of the drugs that would cause slow suffocation and cardiac arrest. Executioners had recorded prisoners’ vital signs, but had not understood what they meant.
California’s execution logs revealed another problem as well: the same six prisoners who continued to breathe did not go into cardiac arrest after injection of the third drug, potassium chloride, which the state’s expert had said would kill within two minutes. Given the massive dose of potassium chloride, how could this possibly be? The answer was one of the “profound physiological derangements” that no one saw coming, at least not until researchers documented it: the bolus dose of sodium thiopental had depressed circulation so dramatically that it blunted the bolus dose of potassium chloride. Prisoners’ hearts raced in response to the potassium chloride, but not enough to induce cardiac arrest, leaving them to die by slow suffocation from the paralytic instead.
The findings from California’s execution logs led a federal court to invalidate the state’s lethal injection protocol in 2006. “The evidence is more than adequate to establish a constitutional violation,” the court stated, noting that it was “impossible to determine with any degree of certainty whether one or more inmates may have been conscious during previous executions or whether there is any reasonable assurance going forward that a given inmate will be adequately anesthetized.” The governor has since declared a moratorium on executions in the state, and it remains in place today.
Looking back, it’s fair to say that for the first 30 years of lethal injection, states used a three-drug protocol without understanding how it actually worked. State experts made claims and stated them with confidence, but what they said didn’t turn out to be true. Sodium thiopental didn’t do what states said it would do, and potassium chloride didn’t do what states said either — largely because no one accounted for the possibility that a bolus dose of the first drug would blunt the bolus dose of the third. States had no idea what their toxic drug combinations would actually do. They were slowly suffocating prisoners to death, and they didn’t have a clue.
Excerpt adapted from Secrets of the Killing State: The Untold Story of Lethal Injection by Corinna Barrett Lain. Copyright © 2025 by New York University. Published by NYU Press.