A self-report postal questionnaire was sent to 5000 (3612 male) drivers randomly selected from the policy holders of an insurance company which insures only motorists who have achieved a maximum 'no-claims bonus'. The questionnaire sought demographic and physical details (sex, age, height, weight and collar size), driving history (number of years, average mileage), accident history (number, number of 'near misses'), frequency of events due to sleepiness (accidents, pulling off the road, falling asleep at the wheel), frequency and severity of snoring and sleep pattern and severity of EDS. Of the 2247 (44.9%) responses received, 1609 (44.5%) were male and 638 (46%) female. Snorers were more likely to report daytime sleepiness than non-snorers and were more likely to have had to pull off the road due to sleepiness. Despite this increased sleepiness there were no significant differences in overall accident rate. These results were obtained from a population with a maximum 'no-claims bonus' which would have excluded any subjects who had already had a recent accident due to sleepiness. Thus the results probably underestimate the problem. It is concluded that snorers tend to be sleepier than non-snorers and are more likely to have modified their driving habits as a result. Snoring and daytime sleepiness should be further investigated as contributory factors to road traffic accidents.