A postal questionnaire survey aimed at exploring the relationship between accidents and daytime sleepiness was sent to 9000 male drivers of which 4621 (51.3%) responded (mean age 47.7, SD 17.1). Drivers provided details of the accidents they had experienced in the last 3 y, and identified those factors, including tiredness, they thought contributed to the accident. In addition, drivers completed the Epworth scale measuring daytime sleepiness and reported whether they had felt close to falling asleep whilst driving during the past 12 mo. Analysis of the data showed that 29% of drivers had felt close to falling asleep at the wheel in the last 12 mo, the probability of which depended on Epworth score, age, occupational group, annual mileage, the proportion of time spent driving on motorways and in built-up areas, how long the driver is prepared to drive before taking a break, and whether the driver is driving a company car or not. Overall, about 7% of accident 'involvements' were associated with tiredness (representing 9-10% of accidents)-a figure which is higher on motorways than on rural roads or built-up roads and higher still in the early hours of the morning. Accident rates of company car drivers and/or those who have felt close to falling asleep at the wheel in the last year are shown to be associated with daytime sleepiness. For example, a company car driver who has felt close to falling asleep at the wheel in the last 12 mo and who scores highly on the Epworth scale has an accident liability which is 70% higher than a similar driver who scores zero on the Epworth scale. Snoring every night increases accident liability by about 30%.