The traditional way of diagnosing obstructive sleep apnea syndrome (OSAS) is all-night polysomnographic recording. A proposed alternative is respiratory polygraphy, a simplified procedure that consists in nighttime monitoring of oxygen saturation, oronasal flow and respiratory movements. Our aim was to evaluate the efficacy of respiratory polygraphy in diagnosing OSAS in comparison with conventional polysomnography. We studied 101 patients (92 men and 9 women) who had undergone polysomnography. An apnea-hypopnea index (AHI) > or = 10 was considered to be the diagnostic criterion for OSAS. To assess the diagnostic validity of respiratory polygraphy we considered that an AHI per hour of recording > or = 10 and a desaturation index per hour of recording > or = 10 were consistent with a diagnosis of OSAS. Sixty patients were diagnosed of OSAS. The AHI per hour of recording was > or = 10 in 56 patients, with 4 false negatives (sensitivity 93.3% and specificity 100%). The desaturation index per hour of recording was > or = 10 in 65 patients, with 7 false positives and 2 false negatives (sensitivity 96.6% and specificity 82.9%). We conclude that respiratory polysomnography is a specific, highly sensitive method for diagnosing OSAS.