To our knowledge, the large spectrum of sleep motor activities (SMA) present in the head and neck region has not yet been systematically estimated in normal and sleep bruxism (SB) subjects. We hypothesized that in the absence of audio-video signal recordings, normal and SB subjects would present a high level of SMA that might confound the scoring specificity of SB. A retrospective analysis of several SMA, including oro-facial activities (OFA) and rhythmic masticatory muscle activities (RMMA), was made from polygraphic and audio-video recordings of 21 normal subjects and 25 SB patients. Sleep motor activities were scored, blind to subject status, from the second night of sleep recordings. Discrimination of OFA included the following types of activities: lip sucking, head movements, chewing-like movements, swallowing, head rubbing and scratching, eye opening and blinking. These were differentiated from RMMA and tooth grinding. The frequency of SMA per hour of sleep was lower in normal subjects in comparison with SB patients (P < 0.001). Up to 85% of all SMA in normal subjects were related to OFA while 30% of SMA in SB patients were related to OFA scoring (P < 0.001). The frequency of RMMA was seven times higher in SB patients than in normal subjects (P < 0.001). Several SMA can be observed in normal and SB subjects. In the absence of audio-video signal recordings, the discrimination of various types of OFA is difficult to achieve and may lead to erroneous estimation of SB-related activities.