BACKGROUND: Many patients with obstructive sleep apnoea (OSA) have difficulty in driving and experience increased automobile accidents. It has previously been shown that patients with OSA perform poorly on a laboratory based divided attention driving test (DADT). METHODS: Seventeen men with OSA of mean (SD) age 49.7 (11.2) years and an initial apnoea/hypopnoea index (AHI) of 73.0 (28.9) were restudied from one to 12 (mean (SD) 9.2 (4.2)) months after initiating treatment with nasal continuous positive airway pressure (CPAP) to examine the effects of treatment on DADT performance. Eighteen age and sex matched controls were also retested 8.4 (3.4) months after their initial tests. Following a practice session, all subjects were given the DADT for 20 minutes before each daytime nap of the standard multiple sleep latency test (MSLT). RESULTS: Untreated patients with OSA, who performed much worse than controls in all measures, improved significantly on all measures of performance, particularly in tracking error which returned to the level of controls in all but one patient. Changes in performance were much greater for patients with OSA than for controls in tracking error (mean difference 106 (95% CI 75 to 135) cm), sleep latency/ MSLT (5.3 (95% CI 2.7 to 8.0) min), number of correct responses (1.2 (95% CI 0.4 to 1.9)), number of missed responses (1.7 (95% CI 0.9 to 2.3)), and number out of bounds (10.0 (95% CI 7.9 to 13.6)), but not for response time (0.1 (95% CI -0.3 to 0.2) s). Improvement in tracking error was highly correlated with improvement in sleepiness (r = 0.65). CONCLUSIONS: Impairment in laboratory driving performance skills in patients with OSA is reversed by successful treatment with nasal CPAP. Changes in daytime sleepiness account for some but not all of the improvement.