A convenient method of monitoring ingestive behaviour in grazing animals is lacking. Acoustic monitoring is a promising technology that could provide accurate and continuous information on ingestive behaviour. A critical step is to develop a signal processing algorithm to automatically classify sound bursts into bite, chew and chew-bite jaw movements. We tested the classification of jaw movements on the basis of six extracted features. Discriminant analysis achieved a correct classification rate of 81.8% using four of the six features. These results suggest that acoustic monitoring may be a simple and effective way of quantifying key aspects of ingestive behaviour in grazing ruminants.