There are very few data available on the effects of respiratory protective masks on sleep. This study was to undertaken to investigate the effects of the M-40 protective mask on sleep architecture. We asked 21 healthy volunteers (18-35 yr) to participate in a two-night study. During the first night, subjects underwent standard polysomnography without the M40 mask. The second night entailed sleeping while wearing the mask, without the hood. Specific variables measured on each night included sleep stage distribution, time awake, number of awakenings, oxygen saturation, number of apneas and hypopneas, and time to sleep onset. Statistical analysis was performed using a two-tailed, paired t-test for all variables except REM sleep, for which a one-tailed, paired t-test was used. There was a significant difference between mask and non-mask nights when measuring total sleep time, total awake time, percentage of time in REM, and number of awakenings. It is concluded that sleep in the M40 mask is disrupted compared to sleep without the mask, with significant decreases in total sleep time and REM sleep.