It's a guilt trip that most of us have made. It's the middle of the night, you've got 200 miles to go, and you're knackered. There's a service station a couple of miles ahead; Knutsford services, however, isn't quite your idea of an enticing place to stop. Besides, the road ahead is empty, and most of the police have knocked off for the evening. Of course you can make it.

However, as is becoming increasingly apparent, your chances of success may be lower than you think. Because it is emerging that the chief cause of road deaths is not booze, drugs, excessive speed or mobile phones; it is good old-fashioned fatigue.

The seriousness of the problem is highlighted by new research carried out at Loughborough University. Using a custom-built driving simulator in the university's sleep laboratory, Professor Jim Horne and his team have been subjecting volunteers to a series of tests to discover how their road-handling skills were affected by lack of sleep.

My experience showed just how risky the game can be. The night before my test I went to bed, as instructed, at 2am. Next morning, at the sleep laboratory, after the prescribed four hours and 50 minutes' sleep, I was festooned in electronic ligaments that feed eye movement and evidence of brain activity back to the monitors. Thus wired, I got into the cabin and began my virtual motorway drive to oblivion.

Just 30 minutes in, I was starting to flag. The technical assistant would periodically shout, "Sleep check!" whereupon I would be expected to give a rating of one to 10 (10 being perky, one being zonked) supplemented by an A to E (A meaning you don't expect to fall asleep in the next five minutes, E indicating you're comatose). Within a few miles, I'd moved from the relative safety of being an 8A to the sleep-infested waters of a 7C. But when I watched the video afterwards, I could see that I was more of a 5D - head leaning back against the headrest in a pathetic attempt to disguise my desire to flop.

I was not the only one to disgrace myself. After my test, Professor Horne showed me the tape of another subject (let's call him Patient X). Head lopped over, eyes rolled back, foot to the floor, Patient X was doing a steady 70. The vibrations from the hard shoulder rumble strip were pulsing through the cabin, but still he slumbered, arms hanging loosely from the steering wheel. After 15 minutes, he stirred, lifted his leaden lids and said he was feeling better. Soon afterwards, however, he was back in a slumber, charging on, eyes wide shut.

Having professed a tendency to doze off, Patient X had been a willing guinea pig for Horne's experiments. But he was so embarrassed by the recorded footage that he insisted that his identity be protected.

Horne's full test goes on for two to three hours and participants take it four times. His "success" rate would have the makers of Mogadon and Classic FM compilations seething with resentment; in the cosy, dark confines of a sawn-off Vauxhall Cavalier, after five hours' sleep, you have only a 10% chance of staying fully awake for a lengthy journey.

The proportion of people who fall asleep is not the only surprising statistic. Forget the image of doddery old ladies nodding off at the traffic lights; the highest risk group is that of male drivers aged under 30. And no matter what your age or gender, driving between 2am and 6am puts a spanner in your body clock, because at this time our bodies are programmed to sleep. Age is more likely to be a factor in the daytime; the older you are, the more susceptible you are to the corollary of the "afternoon dip", the time when most Europeans have wisely gone back to bed. And if you're a student, journalist, nurse, sales rep, commercial driver, doctor, pilot or night-shift worker, you're even more likely to suffer from drowsiness at the wheel.

The annual report on road accidents by the Department of the Environment, Transport and the Regions revealed that 3,500 deaths occurred on Britain's roads in the year to 1998. Of these, police believe, SRVAs, or sleep-related vehicle accidents, claimed more lives than drink-driving. The problem is becoming so familiar to emergency services working on motorways and trunk roads that nicknames have been coined. Part of the M6 is called "the Sleepy Corridor" and the M1 at Boot Hill in Northants has become the "Nod-off Zone".

As part of Professor Horne's work, researcher Diane Flatley is compiling a league table of the UK's most soporific highways, the details of which will land somewhere on a desk in the Department of the Environment, Transport and the Regions this autumn. Top of the list is likely to be the M180, a popular haunt of Grimsby commuters. It has recently been the setting for four fatal crashes.

The SRVA figures are a wake-up call to road designers, many of whom seem to have overlooked the fact that if you drop drivers into a closed loop of repetitive tarmac, things will go wrong. Professor Horne wouldn't disagree - indeed, highways authorities have taken his team's advice and removed many lay-bys on gradual right-hand bends because they are a magnet not just to tired drivers but also to those who have already fallen asleep.

Professor Horne's focus, however, is more on the driver than on the road. He believes the real issue is personal: "People know when they are sleepy, they know if they are reaching the point of fighting sleep. But those who survive sleep-drive accidents will often say, 'I just suddenly fell asleep.' There is no truth in this. You can say, 'There but for the grace of God go I,' but there is simply no excuse for falling asleep at the wheel. We know the signs, but we choose to ignore them."

So what can be done to reduce your chances of succumbing to sleep? "Everyone has their own pet theory," says Horne. "We try all sorts of things, from turning up the radio, opening the window, chewing lemons and taking your shoes off. But there's no escape - if you're tired, you can even fall asleep with your eyes open." With luck, these results may help the DETR to formulate a sleep-drive rule equivalent to the recommendations on drinking and driving.

One of the likely guidelines is that five hours is not enough. Anything less than 300 minutes' sleep gives you Margaret Thatcher eyes and puts you on the cusp of sleep deprivation. Later that day, drive for more than two hours and the chances are that you'll feel the onset of fatigue - as 90% of Horne's guinea pigs did. (If, by the way, you've ever felt your head dip when behind the wheel, you've satisfied the lab definition of nodding off; slack neck muscles signal an advanced stage of sleep.)

The verdict will also nail the truth about coffee. The team has explored its potential to prevent sleepiness, but found that the all-important caffeine levels vary so widely both between and within brands that the results could not be measured. Subsequent tests of coffee at motorway service stations across Britain showed that while you need a shot of 150mg of caffeine to resharpen your senses, many service-station coffees contained no more than 30mg. So you'd need at least five cups of coffee (and a jumbo bladder) before you would be safe to drive again.

Furthermore, no brand of coffee lists its caffeine content, which is why the lab tests on the effect of caffeine were conducted with the help of the makers of the energy drink Red Bull. Two of the slimline cans packs the required 150mg punch; the efficacy of these countermeasures, when combined with a brief nap (and double-blind-tested using a caffeine-free version), looked promising.

"The results are not conclusive, but are becoming more so. So much advice, like stretching your legs and getting fresh air, is a waste of time; if you're tired, the only thing worth doing is sleep - a 15-minute nap is ideal, because it stops short of deep sleep, which can leave you groggy. Take half an hour's break in all to allow time for the caffeine to take effect."

All of which is helpful, but what of the prime victims, males under 30? Why are they so vulnerable to the risk of sleeping at the wheel? There is no conclusive data here, but the team have some inkling. "Fewer women drive at the high-risk times and women are quicker to recognise that they might be sleepy in the next five minutes; men take more risks; they deny being tired and say they can get through it. If they're half an hour from home, they won't stop. Which is why some never make it."

* Plan your journey

Avoid driving for extended periods between the hours of 2am and 5am, and in the early afternoon. Build in breaks at least every two and a half hours.

* Drive refreshed

Aim for at least seven hours' sleep the night before you travel.

* Find a safe place to rest

Whenever you feel tired, get off the road altogether - park away from passing traffic.

* Keep your caffeine levels up

Anything less than 150-200mg of caffeine is unlikely to be of any benefit (energy drinks list their caffeine content; coffees do not). Wait for the caffeine to take effect before you start driving again - rest for half an hour afterwards.

* Don't overdo it

Try not to sleep for more than 15 minutes to avoid the grogginess that can result from deep sleep.