We studied male patients (BMI = 27.6 +/- 3.4, mean +/- SD), mean age 54.1 +/- 8.9 years, with stable NYHA class 3-4 congestive heart failure (CHF) (LVEF = 24.3 +/- 11.5 percent) and normal daytime arterial blood gas values. These patients underwent three consecutive nights of full polysomnography; adaptation, control, and treatment with nasal CPAP. Each night's study was followed during the day by cognitive testing and multiple sleep latency tests (MSLT). The purpose of the study was to document the effect of nasal CPAP on these variables. The main findings of the study showed no significant differences between control and treatment nights with respect to the amount of Cheyne-Stokes respiration (CSR) observed, the nocturnal oxygenation, or sleep quality. Both subjective and objective measures of sleep quality showed no change from night to night. In addition, the degree of cognitive functioning and daytime sleepiness (as measured by MSLT) showed no significant differences between control and treatment nights. We conclude that short-term treatment with nasal CPAP in patients with CHF does not improve either CSR, nocturnal oxygenation, or sleep quality. Furthermore, most of our patients did not tolerate nasal CPAP therapy.