Recent objective studies demonstrate relatively low hours of nightly use during nasal continuous positive airway pressure (CPAP) therapy for obstructive sleep apnea (OSA). Patients frequently complain of dyspnea or discomfort during CPAP use, especially during expiration (against the continuous pressure), which may be a reason for the low hours of use. We hypothesized that with decreased expiratory pressure, hours of nightly use would increase. Therefore, we randomized 83 OSA patients to receive either continuous or bilevel positive airway pressure when expiratory pressure is lower. To document objectively the effective use of either therapy, we built and installed elapsed-time and mask pressure sensors in the patients' positive airway pressure units. A total of 62 patients were evaluable and followed for 1 yr. Of these, 26 received bilevel and 36 CPAP pressures. The machine timers measured accumulated "machine-on" time, and the mask pressure sensor recorded the total time in which the mask pressure was within 2 cm H2O of the effective pressure (pressure shown to eliminate 95% of the obstructive apneas during a full night of polysomnography). The mean machine timer hours of CPAP were 5.0 +/- 0.19 SEM and 4.9 +/- 0.23 SEM during bilevel therapy (p NS) over a 12-mo period. The pressures required during CPAP or bilevel therapy were not different between high and low hourly users. Effective use, the percentage of time that the machine was running and the prescribed pressure was being delivered, was 80% in CPAP and 82% in the bilevel users (p NS). Both groups had equal complaints with regard to mask discomfort, machine noise, and nasal stuffiness.(ABSTRACT TRUNCATED AT 250 WORDS)