Estimating the size of badger populations at a local scale using direct observation is difficult because of their nocturnal and elusive lifestyle. Methods based on live-trapping can provide robust estimates of badger numbers, but are very labour-intensive and expensive. The development of alternative methods to predict badger numbers accurately would provide valuable tools for research and management. Badger setts are distinct structures, which exhibit various signs of activity. The physical appearance of setts has been used in the past as a crude guide to the number of badgers likely to be trapped. In this study, an objective system for quantifying levels of activity at badger setts was developed and related to the number of trapped badgers. Several types of activity were positively associated with badger numbers, although these associations varied seasonally. Furthermore, all individual activity scores, summary indices and combinations of activity types were poor predictors of the number of badgers trapped. This is probably because variations in environmental conditions and badger behaviour confound the relationship between activity and badger numbers.