Characterization of settlement patterns is one of the core concepts in archeological research. The duration of an occupation is usually estimated through zooarchaeology (e.g., density of remains, cementochronology) and is limited by taphonomic processes and sample size. We propose a new application of dental wear methods for estimating the relative duration of hominid settlements in Paleolithic sites. Dental microwear is known to be sensitive to seasonal changes in diet. In this new application we use microwear scratch counts to estimate the variation in the dietary signal of various ungulate species. We propose that this variation is correlated to the duration of site occupation. Each season presents a limited and different set of food resources available in the environment. If animals are sampled only during a specific season (i.e., during a short term occupation) then they would be expected to have a dental wear signal with little variation. On the other hand, a greater diversity of food is available across different seasons. Therefore, if game animals are hunted through various seasons during long occupation periods, then they would be expected to have more variable dental wear. The application of this technique to the Middle Paleolithic site of Arago Cave (France), where various types of occupations occurred, supports this hypothesis. When combined with multidisciplinary studies of archaeological localities (seasonality in particular), this new application of dental wear analysis presents valuable information about hominid settlements and behavior. We contextualize our data with results from lithic and zooarchaeological analyses from Arago. These results reveal the presence of both high and low mobility groups of Homo heidelbergensis throughout the sequence of the Arago Cave.