An empirical method for investigating differences in neural control of jaw movement across oromandibular behaviours is to compute the coherence function for electromyographic signals obtained from mandibular muscle groups. This procedure has been used with adults but not extended to children. This pilot study investigated if coherence analysis could reveal task-related differences in control for children by measuring mandibular electromyograms obtained from an infant and adult. Electromyographic signals were obtained from bilateral masseter and temporalis muscle groups during chewing and babble from a typically developing infant from 8-22 months, and chewing and speech were obtained from an adult. Coherence functions were computed. Measures obtained from the infant and adult exhibited a significant main effect for task, with peak coherence values within 20-60 Hz being significantly greater for chewing than vocalization. This pilot study suggests that coherence analysis of mandibular EMG is a sensitive measure for distinguishing task-related differences in neural organization for children. (Contains 2 tables and 5 figures.)