During 1987 through 1992, 85 wolves (Canis lupus) were captured, radio-collared, and relocated from aircraft 1,123 times in northwest Alaska. Wolf packs usually did not follow migratory caribou (Rangifer tarandus) but maintained year-round resident territories that averaged 3,652 km². During years when caribou were absent and moose densities were low, ≤ 25% of the wolf packs moved 64 to 272 km to the caribou wintering grounds. Wolves used different slopes, aspects, and habitats in summer versus winter. Twenty-five percent of the radio-collared wolves dispersed. Annual finite rates of increase ranged from 0.64 to 1.43. Annual wolf survival rates averaged 0.59. There were differences in survival rates among years. Sixty-one percent of the wolves died. Hunting was the main cause of death (69%) followed by rabies (21%). Rabies was a significant natural limiting factor. This wolf population could sustain mortality rates of about 53% annually. Caribou and moose composed 51 and 42%, respectively, of the observed wolf prey. Adjusted for prey size, each pack killed 1 adult moose equivalent per 6.7 days. Wolf pack sizes and adjusted kill rates and kgs of available prey per wolf per day were correlated from several areas across North America. When caribou were present they were the principal prey. However, when caribou densities were <100/1,000 km² wolves preyed upon moose. Wolves preyed upon relatively healthy caribou and moose that were in marginal condition. Wolves were killing about 6-7% of the caribou herd and from 11 to 14% of the moose population annually. Existing wolf predation may have serious impacts on resident, low-density moose populations. During spring 1990 I tested the line-intercept method of sampling tracks for estimating wolf densities for a known wolf population (i.e., 48 wolves). The population estimate based upon line-intercept sampling was 50.7 (80% CI = 33.4 to 67.9) suggesting that the survey method provided relatively accurate population estimates. I placed 23 satellite transmitters on wolves aged 10-months to 8 years with no apparent adverse effects on them. Accuracy of 1,855 relocations at 9 sites averaged 336 and 728 m for best and worst quality relocations, respectively. Satellite telemetry has potential for providing improved data sets for evaluation of wolf territory sizes and movements.