Neurologists need to recognize, diagnose, and treat obstructive sleep apnea (OSA) in patients with stroke or transient ischemic attack (TIA). Increasing medical evidence suggests that OSA is an independent risk factor for stroke and TIA. Stroke (or TIA) is more likely a cause, rather than a consequence, of OSA because PSG studies have shown: 1) apneas in stroke are typically obstructive, not central or Cheyne-Stokes in type; 2) apneas are just as frequent and severe in patients with either TIA or stroke; 3) OSA severity is not influenced by the acuteness or location of the stroke; 4) untreated OSA patients have more strokes, stroke morbidity, and mortality than those who are treated. OSA alone can induce hypertension, especially in younger men. A causal relationship has recently been demonstrated between OSA and hypertension. A distinctive feature of OSA-induced hypertension is loss of the normal nighttime fall in blood pressure (“nondippers”). Data from the Sleep Heart Health Study showed a dose-response association between OSA severity and the presence of hypertension 4 years later. Hypertension or ischemic heart disease usually develops in untreated patients with OSA over time without particular worsening of OSA. Studies have shown sleep itself is not a risk factor for stroke because most stroke and TIAs begin between 6 am and noon, while the individual is awake. However, OSA promptly be considered in stroke beginning during sleep because 88% of strokes that develop during sleep occur in “nondippers.” Premature death in OSA patients is most often cardiovascular, but occurs while the patients are awake. The risk of myocardial infarction is increased 20-fold in untreated OSA. Treating OSA patients with continuous positive airway pressure can prevent or improve hypertension, reduce abnormal elevations of inflammatory cytokines and adhesion molecules, reduce excessive sympathetic tone, avoid increased vascular oxidative stress, reverse coagulation abnormalities, and reduce leptin levels. If all this can be achieved by a polysomnogram, then this test should become part of a neurologist's armamentarium for stroke and TIA.