Previous reports have shown an association between snoring and stroke but it is not clear whether this reflects confounding factors nor whether the association is attributable to obstructive sleep apnoea (OSA). We performed a case-control study of 181 patients admitted to hospital with first-ever stroke and community control subjects matched individually for age, sex and general practitioner. Subjects were interviewed with a structured questionnaire to identify snoring, daytime sleepiness and stroke risk factors. The association between snoring alone and stroke was not statistically significant: odds ratio (95% CI) 1.44 (0.88, 2.41). Daytime sleepiness was, however, significantly associated with stroke: odds ratio 3.07 (1.65, 6.08). Multiple logistic regression showed that hypertension, current smoking, taking alcohol regularly (negatively) and a higher Epworth sleepiness score were independently associated with stroke. The results suggest that the previously reported association between 'simple' snoring and stroke might have been due to poor controlling for confounding variables. Our study suggests an association with greater sleepiness prestroke, the cause of which is unclear, although OSA is a possible candidate.