Bear food habits are often quantified using scat analysis, mainly due to its non-invasiveness and because samples are relatively easy to collect. However, lab processing time can be daunting and may end up competing with other field activities. Sub-sampling a bear scat to analyze its contents may reduce the lab processing time, but the number of subsamples per scat is usually chosen arbitrarily. We investigated the effect of the number of subsamples per bear scat on the estimatation of the diet composition of the Apennine brown bear in the Abruzzo Lazio and Molise National Park. Based on a sample of 328 bear scats collected in 2006, and from 5 to 1 subsamples (10ml) per scat, dietary analysis showed qualitative and quantitative stability at a decreasing number of subsamples, and only food items of negligible importance were occasionally missed using 1–2 subsamples per scat. We concluded that 2 subsamples can be used without significant loss in accuracy, corresponding to a 60% reduction in lab time, and to more than 50days of lab work for one operator to process our entire bear scat sample. By assessing the effect of sub-sampling a bear scat for dietary analysis, we also present preliminary data on the seasonal food habits of the Apennine brown bear population.