Since sleep apnea (SA) and stroke have many shared risk factors an independent contribution of SA to the overall risk of stroke is not easily proven and has been questioned recently. To contribute to this controversy, we analysed the frequency of SA in groups of patients with first and recurring ischemic stroke. We prospectively studied 102 patients admitted to our stroke unit. The prevalence of vascular risk factors and a history of previous stroke were recorded. All patients received cardio-respiratory polygraphy during the first 72 hours after admission. CT and MRI scans were evaluated for the location of the acute stroke and the presence of older vascular lesions. Thirty-four women and 68 men with a mean age of 64.5 +/- 13.7 years were included in the study. Cerebral lesions attributable to a previous stroke were identified in 25 patients, of whom 19 reported to have suffered a stroke before. Patients with stroke recurrence had a higher mean apnea-hypopnea index (AHI) (26.6/h vs. 15.1/h, p<0.05) and more often presented with a sleep apnea syndrome (SA) defined by an AHI >or=10/h (80 vs. 52%, p < 0.05) than patients with first ever stroke. Logistic regression analysis including the variables "age", "gender", "cumulative risk factors", "AHI >or=10/h", and "diabetes" identified diabetes (Odd's ratio [OR]=4.5) and AHI >or=10/h (OR=3.5) as independent risk-factors for stroke recurrence. According to our results SA is an independent risk factor for stroke recurrence. We therefore advocate routine sleep-apnea screening in all patients having suffered an ischemic stroke.