Observations of wolves on Isle Royale are reported for 1961–66, with interpretations including the earlier 3-year period described by Mech (1966). On this 210-square-mile island the fully protected wolf population varied from approximately 22 to 28 in midwinter. The major and minor foods were moose and beaver, respectively. The main pack varied in number between 11 and 22 with about three breeding pairs believed present. The population remained relatively stable; mating occurred every winter; and adult mortality appeared to be low. High mortality among pups seemed to be the point of population control. Socio-economic factors may have controlled the size of the large pack. Availability of food during the period of parturition and rearing probably was critical to survival of young. Recruitment of young appeared to take place in years of high production of moose calves. Numbers in the large pack probably were curtailed through the progressive exclusion of aged and socially subordinate individuals. Under harassment these animals separated and became pack-following scavengers, then probably true loners ranging outside the area used by the pack. Smaller aggregations of two or three non-breeders were seen each winter, as were the loners, some of which appeared thin and weak. The only known breeding outside the big pack was in a group of five present in the winter of 1965. This group was probably a family unit which separated from the main pack. A year later the male had disappeared, and remains of a pup, probably theirs, were found. In the winter of 1966 the alpha male of the large pack became lame and apparently was killed. This background appears favorable for further changes in social organization.