The conventional study of obstructive sleep apnea syndrome (OSAS) requires two polysomnographic recordings on two different nights, one for diagnosis and the other for determining an effective level of continuous positive airway pressure (CPAP). In this study we examined an alternative method requiring a single night of polysomnographic recording encompassing both diagnostic and treatment functions. In order to obtain recordings of all phases of sleep, as well as recordings taken in supine decubitus position, the duration of polysomnographic readings was individualized ("made to order" polysomnograms). A single-night session was sufficient for diagnosis and for adjusting treatment in most cases. In some cases the night was extended well into the morning hours in an effort to obtain all the necessary data. "Made-to-order" polysomnograms were compared to traditional ones by contrasting breathing parameters (the U of Mann Whitney) and sleep architecture findings for two groups of men diagnosed consecutively of OSAS. Thirty-four patients were studied over two nights (mean age: 50 years; BMI 37 Kg/m2). Forty-six patients (mean age: 49 years; BMI: 38 Kg/m2) were studied in single-night sessions. The differences between the two groups were not significant. For most patients studied in "made to order" sessions, all sleep stages were recorded and readings were taken in supine position. REM sleep was not recorded, however, for 26% of the patients in this group. No significant differences between the double- and single-session groups were found for apnea (51 versus 62), maximum oxygen desaturation (72% versus 75%) and effective CPAP level (9 versus 10 cm H2O). These data suggest that both study methods are equally useful for evaluating OSAS severity and for prescribing effective CPAP treatment. We discuss, however, the need to arrange for two-night studies in the minority of patients for whom no REM sleep recordings become available during a single session.