The aim of this study was to determine if is it required to use a new index to assess the severity of obstructive sleep apnea syndrome (OSAS). 20 OSAS patients (Group I) of whom Apnea-Hypopnea Index (AHI) and clinical findings were evaluated to be discrepant, and 40 OSAS patients (Group II), which were found not to sleep in the supine posture after polysomnography (PSG), underwent a second PSG. The first and the second PSG results of the two patient groups were compared. In the comparison of the two PSG results of Group I, a significant difference was detected between the mean AHI, supine time and the non-supine time. A significant difference was also found between the two PSG results of Group II with regard to the mean AHI, supine time and non-supine time. We concluded that the most effective factor on AHI is the patients’ sleeping time in the supine posture. We offer the use of a novel index (corrected AHI) for evaluating the severity of the disease. We find it useful to perform a second PSG on patients who had not slept in the supine posture or when the clinical findings and PSG data are discrepant.