Objectives/Hypothesis: We used a new automatic snoring detection and analysis system to monitor snoring during full-night polysomnography to assess whether the acoustic characteristics of snores differ in relation to the apnea-hypopnea index (AHI) and to classify subjects according to their AHI Study Design: Individual Case-Control Study. Methods: Thirty-seven snorers (12 females and 25 males, ages 40-65 years; body mass index (BMI), 29.65 +/- 4.7 kg/m(2)) participated Subjects were divided into three groups: G1 (AHI <5), G2 (AHI >= 5, <15) and G3 (AHI >= 15) Snore and breathing sounds were : recorded with a tracheal microphone throughout 6 hours of nighttime polysomnography The snoring episodes identified were automatically and continuously analyzed with a previously trained 2-layer feed-forward neural network. Snore number, average intensity, and power spectral density parameters were computed for every subject and compared among AHI groups. Subjects were classified using different AHI thresholds by means of a logistic regression model. Results: There were significant differences in supine position between G1 and G3 in sound intensity, number of snores; standard deviation of the spectrum, power ratio in bands 0-500, 100-500, and 0-800 Hz, and the symmetry coefficient (P < .03); Patients were classified with thresholds AHI = 5 and AHI = 15 with a sensitivity (specificity) of 87% (71%) and 80% (90%), respectively. Conclusions: A new system for automatic monitoring and analysis of snores during the night is presented. Sound intensity and several snore frequency parameters allow differentiation of snorers according to obstructive sleep apnea syndrome severity (OSAS). Automatic snore intensity and frequency monitoring and analysis could be a promising tool for screening OSAS patients, significantly improving the managing of this pathology.