The study of sleep bruxism is usually based on clinical history, signs and symptoms. The recording of electromyographic signals with either ambulatory portable home recorders or with polysomnographic techniques in the laboratory environment allows collection of objective data. The present study showed a 100% agreement with clinical evaluation in the recognition of bruxism episodes from the masseter electromyogram recorded with portable recorders and using the polysomnographic technique. On the contrary, scorers had difficulties in discriminating between different types of episodes (phasic, tonic and mixed), the between-scorers agreement varied between 62% and 63% and the kappa-values between 0.43 and 0.33. The ideal time base at which electromyographic signals should be integrated to allow for a good discrimination of bruxism patterns is 0.06 s. The results indicate that portable electromyography recorders are a valuable complement to polysomnographic recordings of orofacial motor activities as they provide a very good recognition rate with adequate time base data collection.