Obstructive sleep apnoea (OSA) is an independent risk factor for hypertension, which is a major cause of stroke. The prevalence and associations of OSA in a cohort of stroke patients were studied. The safety and tolerability of early treatment with nasal continuous airways pressure (nCPAP) was also assessed. Consecutive subjects admitted with acute stroke were assessed clinically, radiologically and with scales assessing prior OSA risk, dysphagia and disability. Sleep studies were performed within the first few days of admission using a portable diagnostic system. Twenty-nine of 55 (53%) subjects had evidence of OSA, using an apnoea-hypopnoea index (AHI) of 10 or greater. The AHI was significantly associated with an index of prior OSA symptoms, but not with history of hypertension, degree of dysphagia, or type and severity of stroke. Use of a portable diagnostic system for detecting OSA in the acute stroke setting was well tolerated. OSA is common after acute stroke and exceeds rates seen in control populations of similar age (53% vs. 11%). Early treatment with nCPAP was effective and well tolerated.