Ask someone to plunge their forearm into a bucket of icy water until the pain becomes intolerable, and they might last just five seconds. Ask someone else, and they might happily go for three minutes. It is no secret that people's perception of pain is as different as their taste in clothes, says Mark Hutchinson of the University of Adelaide, Australia. But Hutchinson thinks he just might have come up with the first objective test of that subjective perception - a development that could revolutionise pain treatment. How the test actually works remains a mystery. But it does suggest that traditional theories to explain the experience of pain are telling only part of the story.

It has long been known that the immune system is linked to pain. In response to infection or a wound, certain immune cells can cause inflammation, which itself is often painful. But there are new, intriguing hints that other sorts of immune-like cells that are found packed around neurons and nerves might in fact have a large part to play in the level of pain we feel when we're exposed to a trigger like a dog bite, surgery - or cold water. These cells share most of the properties of the standard immune cells that circulate in the blood.

The conventional idea that nerves transmit pain signals to the brain is true. "But there is animal research that suggests these immune-like cells in the spinal cord can act a bit like a dimmer switch to control how much of a pain signal is sent by the nerves," Hutchinson says.

The Adelaide research involved scoring healthy people for levels of pain tolerance, using the cold water test. Then, Hutchinson took blood samples, exposed those samples to a compound that triggers immune cells to replicate, and stirred morphine into the mix.

When he added a high dose of morphine to the sample (enough to kill someone if their bloodstream were to reach the same concentration) the rate at which the blood immune cells replicated mapped directly on to the amount of time they had lasted in the cold water test. "We did some more work to check this wasn't some kind of artefact, but, whammo, the two scores kept going into that beautiful line," he says.

The best current objective tests for pain tolerance, based on a variety of markers, such as blood biochemistry and genetics, are only about 20-30% accurate. "This was a small, initial study - but in these 13 healthy people, we had upwards of 90% accuracy," says Hutchinson. It is early days, and the test would need to be tried out in a much larger number of people, and sped up - it now takes 24 hours - but in theory, it could be hugely valuable.

"If you had the ability to predict how sensitive someone was to pain, it would revolutionise how you treated pain in that person," says Hutchinson. Since some people can last five seconds at the cold water test and some three minutes, the dose of painkiller they'd need for an injury being treated in the emergency room, or for childbirth or surgery, would vary hugely.

"At the moment, a lot of people are under-dosed, so they're still experiencing pain, but other people are getting too much medication, which, because their bodies are over-responding to the 'standard' dose, could get to toxic levels," he says.

Dr Peter Cabot, a pain researcher at the University of Queensland, agrees that if the test is validated by trying it out on many more people and different groups, such as the very old and very young, it could be hugely beneficial, especially for treating people who cannot communicate their pain effectively, like children, patients with dementia or heavily sedated burns victims. "Then, an objective test would be hugely valuable," he says.

While Hutchinson is excited about the potential for a pain test, he is even more interested in how it might actually work. The immune cells isolated from blood are very unlikely to have any impact on pain sensation. It is those specialised immune-like cells in the central nervous system that would have to be involved.

So perhaps the test works because all these blood and central nervous system immune cells share some characteristics, and the morphine is somehow tapping into these, he reasons. But exactly what those characteristics might be, and exactly what determines pain tolerance threshold, are questions still to be answered.

"That's the exciting thing," Hutchinson says. "Every day, we go to work and go exploring into pain where no one has gone before."