During 1978-81, 38 male and 55 female caribou (Rangifer tarandus) were successfully radio-collared in the range of the WAH and monitored on a year-round basis for 322 and 736 collar-months, respectively. Males were relocated 83 times and females 279 times. Collared males died or shed collars more frequently than females. As of 31 December 1981, radio collars were functioning on a maximum of 19 males and 53 females. By relocating collared caribou for 2 or more years, we found that the only movement patterns of collared individuals included the following (1) females returned each May to the traditional WAH calving area on the North Slope; (2) both sexes used the foothills near the calving grounds in May and June; and (3) both sexes returned to the arctic coastal plain in summer. Use of a particular winter range during successive winters by an individual was not predictable. In addition, the number of caribou using a particular winter range varied greatly between years, although the most important wintering areas consistently included the Selawik and Buckland River drainages, the arctic coastal plain, and the central Brooks Range. Because of the current and historic seasonal distribution patterns of the region's caribou, policies for managing NPS lands in northwestern Alaska will generally have substantial effect on important direct mortality factors to caribou. These policies will affect the caribou population by controlling hunting, both for subsistence and sport, and predation through protective regulations for major predator species. NPS policies will have less impact on welfare factors such as habitat because critical seasonal habitats extent well beyond or occur totally outside NPS lands. The most unexpected aspect of habitat selection by the WAH, and the one possibly having greatest habitat management implications, was that spruce (Picea spp.) forest habitat was rarely used even in winter. Instead, open tundra plateaus, tussock communities, and windblown ridgetops were favored. Food habits of caribou wintering on the arctic coastal plain apparently differed substantially from those of caribou wintering in the Selawik Hills and south of the Brooks Range. For instance, caribou feces collected on the arctic coastal plain in late winter contained about 40-50 percent less lichen fragments and 4 times more shrub fragments (primarily Vaccinium) than similar samples collected south of the Brooks Range. Review and assessment of effects of fire on the WAH suggest to us that wildfires had little or no effect on the WAH's decline between 1970 and 1976. The WAH apparently uses tundra wintering habitats more than tree-dominated winter habitats. Consequently, emphasizing research on tundra fire ecology rather than taiga fire ecology should be more relevant to managing the WAH.