The nociceptive trigeminal inhibitory (NTI) splint has been claimed to decrease the electromyographic (EMG) activity of jaw-closing muscles and relieve symptoms of various types of temporomandibular disorders (TMD) and bruxism. The present study was designed to address the question about EMG-changes during sleep. Ten patients (age: 23–39 years) with a self-report of tooth-grinding during sleep were recruited. Patients were examined at baseline and after each treatment period with the use of the Research Diagnostic Criteria for TMD. A portable EMG-device was used to record EMG-activity from the masseter muscle during sleep. The patients received two 2-week splint treatments in a randomized cross-over fashion; an NTI splint and a standard flat occlusal splint (OS). EMG data were analysed according to published criteria. Using a 10% of maximum clenching EMG-activity cut-off threshold to determine the number of EMG-events h−1 of sleep, the NTI splint was associated with a significant reduction (9·2 ± 3·2 events h−1) compared with baseline EMG (19·3 ± 4·0; ANOVA: P = 0·004, Tukey post hoc: P = 0·006), whereas there were no differences between the OS (16·2 ± 4·7) and baseline EMG (19·2 ± 4·1; P = 0·716). There were no effects of either NTI or OS on clinical outcome measures (ANOVAs: P > 0·194). This short-term study indicated a strong inhibitory effect on EMG-activity in jaw closing muscles during sleep of the NTI, but not the OS. However, the EMG-activity was not directly related to clinical outcome. Further studies will be needed to determine long-term effects and possible side effects of the NTI splint.