A cable tie, a carabiner, and a boy’s hands fastened behind his back. That was the last step before a rescue diver rolled him face down into black water and swam him out of a mountain.

Thirteen people went into Tham Luang Nang Non on 23 June 2018 and none of them walked out. Twelve boys from the Wild Boars football side, aged 11 to 16, and their 25-year-old coach went in after training, and early monsoon rain sealed the entrance behind them. British divers Rick Stanton and John Volanthen found them alive nine days later, four kilometres inside, huddled on a muddy shelf above the water. Stanton told reporters in Chiang Rai that he counted the figures coming round the rock face until he reached 13, and called the relief massive.

Every other option had been eliminated

Finding them turned out to be the smaller half of the problem.

Richard Harris, an Adelaide anaesthetist and 30-year cave diver, got the call on 5 July and flew in that day with his dive partner Craig Challen. By then the alternatives had been worked through and binned. Pumping was losing to the rain, and drilling from the mountain above was guesswork. Waiting out the monsoon meant months underground. Teaching thirteen non-swimmers to dive a route with six flooded sections, moderate current, near-zero visibility and line traps (gaps a guide rope slips through but a diver cannot) risked panic underwater, which could drown the boy, the diver, or both.

Urgency was not theoretical. Saman Kunan, a 38-year-old former Thai navy SEAL, died on 6 July after laying air cylinders along the route, as CBS News reported at the time. Oxygen where the boys were sitting had dropped to about 15 per cent against a normal 21, authorities confirmed at the time.

Why the drug was ketamine

Harris canvassed paediatric anaesthetists and psychiatrists and weighed benzodiazepines, clonidine and chloral hydrate before landing on ketamine. Writing later in Diving and Hyperbaric Medicine, he and his co-authors flagged a gap in the literature that should worry anyone: apart from Edgar Pask’s wartime work, no precedent existed for putting an anaesthetised human underwater.

Ketamine won on unglamorous grounds. Injected into a muscle, it needs no drip line. People under it keep breathing on their own and hold on to most of their airway reflexes. There is a wide gap between a working dose and a dangerous one, which matters enormously when the person giving the next injection is a cave diver with an afternoon’s training, floating in a stream passage. And it nudges blood pressure up rather than down, handy when no monitoring exists.

Sorted into small and large

How do you dose someone correctly when most of the people injecting them have never done it before? Weights ranged from 30 to 65 kilograms, so everyone was simply filed as small (roughly 40 kilograms) or large (roughly 50), with 45 as the cutoff. Each boy fasted from 6am, took half a milligram of alprazolam by mouth, was dressed in a wetsuit and a horse-collar buoyancy device, then walked down a 15-metre mud slope to Harris, who put atropine into one thigh and ketamine into the other, straight through the neoprene with a 23-gauge needle. Top-ups were pre-drawn at 100 milligrams for a small and 125 for a large. Each rescuee needed three or four, given by divers in the dry chambers between sumps.

Mark Irwin, a bioethicist at the University of Otago, went back over the ethics of all this in the Journal of Bioethical Inquiry and noted the training those divers received: a short talk from Harris and practice injections into a plastic bottle, the day before.

How the mask kept water out

Before the first boy went under, Harris and the diver checked the mask’s seal three times, sitting him upright between checks to look for water.

Keeping an unconscious person alive underwater is normally considered a losing bet, so diving guidance says get them to the surface immediately. Harris had no surface to go to. What he had was an Interspiro Divator full face mask with a safety pressure function, keeping the inside at higher pressure than the water outside so leaks push out rather than in. Straps were tightened past the point of comfort. A cylinder of 80 per cent oxygen went on the chest. Exhaled bubbles were the only monitor, and when a boy stopped breathing, a diver pressed the purge valve and gave him a shove of pressure to start again.

It held. At the field hospital, the full-face masks came off for nonrebreather oxygen masks and the wetsuits were removed carefully with the neck supported, according to the medical team’s research letter in the New England Journal of Medicine. Twelve of the thirteen needed nothing beyond that and supplemental oxygen, the same team reported in a longer paper published the following year. One boy had a brief laryngospasm. Three had chest X-ray changes consistent with minor aspiration or infection. One arrived at 29.6 degrees, severely hypothermic and not even shivering, was rewarmed, and was fine. None of them remembered a thing.

One case is one case

That same paper included a bench test: the University of Auckland team put the mask in a pool, confirming it does maintain positive pressure on open-circuit scuba. They still declined to rewrite the rulebook, concluding that the safest place for an unconscious diver remains the surface and that Thailand’s success rested on a stack of conditions unlikely to line up twice: lean children, expert divers, shallow water and days to plan.

Harris has never framed it as a triumph of method. Talking to Deadline about the documentary made of the rescue, he said he’d been certain it wouldn’t work. He went in anyway, and it worked.

Correction, 15 September 2026: The excerpt was revised to clarify that the team was found after nine days and fully rescued 17 days after entering the cave, rather than being trapped for only nine days.