Analysis of short-term feeding behaviour may improve our understanding of food intake regulation and diet choice. Feeding behaviour of animals consists of feeding events separated by non-feeding intervals. Feeding events are often observed to be clustered into bouts that may be called meals. Determining a meal criterion (the longest non-feeding interval which is accepted as part of a meal) allows feeding events to be grouped into meals. Tolkamp & Kyriazakis (1999) presented a model that described three populations of loge-transformed intervals in the form of three normal distributions (Gaussians). These populations represent intervals within meals, with or without drinking, and intervals between meals. This model predicted that the probability of a meal starting, first increased, then decreased, with time since the last meal. This is in contrast to the satiety concept, which predicts that the probability of an animal starting a meal will increase with time since the last meal. This study aims at developing a model that best describes, biologically and statistically, the distribution of non-feeding intervals, thus leading to a more accurate meal criteria.