Patients with sleep apnea are typically hypersomnolent during the daytime and may demonstrate higher order cognitive dysfunction. A persistent problem in interpreting impaired neuropsychological test performance in such patients is whether the observed deficits can be explained wholly by impaired vigilance. We examined 37 sleep apnea patients prior to and immediately subsequent to successful sleep apnea treatment with nasal continuous positive airway pressure (CPAP). Patients were evaluated immediately after morning awakening in the sleep lab. A brief neuropsychological evaluation, was administered at that time. Following this, alertness was measured with a 30-min polysomnographically determined sleep latency test. Both test (approximately 50 min in duration) were performed once following a baseline (diagnostic) night in the sleep lab and once in the morning following a CPAP (therapeutic) night in the lab. Subgroup analyses indicted that while vigilance impairment can account for some of the decreased test performance seen in sleep apnea (auditory verbal learning) the effects of severe nocturnal hypoxemia appear to affect other function (sustained attention in repetitive arithmetic calculations) that were not easily modified by treatment. Thus, performance on the recall trial of the Rey Auditory Verbal Learning Test increased from pre-CPAP to post-CPAP for the increased alertness group but decreased significantly for the decreased alertness group. On the Wilkinson Addition Test, non-hypoxemic patients showed statistically significant improvement in problems correctly solved from pre-CPAP to post-CPAP, but the hypoxemic patients showed only a marginal increase. These results are compatible with other studies suggesting that patients having sleep apnea may incur deficits as a result of both decreased vigilance and hypoxemia, and that at least some of these deficits are immediately reversible.