All-night polysomnographic recordings were made of clinically diagnosed sleep bruxists (n = 23) and non-symptomatic controls (n = 6). The total duration of masseter contraction (MC) episodes during sleep was 11.6 minutes per night in bruxists and 6.6 in controls (P < 0.01). The mean frequency of MC episodes was 11.0 per hour of sleep in bruxists and 6.4 in controls (P < 0.05). The mean relative amplitude of MC episodes reflecting clenching strength was 0.81 in bruxists and 0.56 in controls (P < 0.01). The percentage distribution of mixed and phasic MC episodes was 94% among bruxists and 88% in controls. The remaining activity was classified as tonic in both groups. The subclassification of rhythmic jaw movements (RJM), defined as three or more separate rhythmic contractions during MC episode were also evaluated. The frequency of those MC episodes with RJM was 3.6 per hour of sleep in bruxists and 1.1 in controls (P < 0.001). The difference in the relative amplitude between the two study groups suggests that the amount of clenching force is the primary factor responsible for the harmful effects of sleep bruxism on the masticatory apparatus. The most significant difference was seen in those phasic and mixed MC episodes which also included the subclassification of rhythmic jaw movement. The result suggests that relative amplitude and rhythmicity of MC episodes can be used as a basis to confirm the diagnosis and to evaluate the treatment effects of suspected sleep bruxists.