A sample of 166 Stoats collected from Craigieburn Forest Park, Canterbury, New Zealand, was used to assess the usefulness of seven different methods of age determination. All the methods made use of characters which do change with age, but not all are equally good at defining useful age classes. The recommended approach is to use a combination of skull and baculum measurements to identify young animals, followed by counting of the canine cementum annuli of adults. If the skull or baculum is broken or not available, visual assessment of the status of the nasal sutures, the lateral suprasesamoid tubercle of the femur and the wear of the carnassial teeth are the next best options for distinguishing the young. Kopein's index based on the closure of the pulp cavity of the canines, and the zonation visible in the periosteal bone of the mandible, are unreliable.