from the Somte montage but these were not clinically important. The Somte significantly underestimated the number of arousals but the number of respiratory events was the same. The positive predictive value for the Somte to detect RDI > 15 was 100% and for RDI > 30 was 80%. Conclusions: In an unselected sleep laboratory population, in the hands of experienced scorers, the Somte accurately measured sleep and respiratory parameters despite the limited EEG montage. Arousals were harder to detect and were underestimated on the Somte . In a laboratory setting the accuracy of the Somte modified polysomnography device was considered adequate for clinical purposes.