A grizzly bear habitat classification was developed by modifying and expanding the climax-based Biogeoclimatic Ecosystem Classification (BEC) to accommodate serai vegetation. Locations of radio-collared bears were assigned to a large number (N=110) of structurally and floristically unique habitats. An interpretive classification of 14 Bear Habitat Units (BHUs) was derived from the taxa of the BEC system; units were amalgamated on the basis of grizzly bear habitat value and similarity of response to forest management practices. Fifteen climax forest, three subalpine, three wetland, and three avalanche chute units were identified and described in the lower Kimsquit River. Two adult female grizzly bears (numbers 08 and 25) were monitored for 1238 and 1196 days, respectively, from April 1982 to October 1985 and had multi-annual minimum convex polygon (MCP) home ranges of 85 km² (N=23 6), and 60 km² (N=241). River floodplain BHUs were used most heavily by bears 08 and 25 during their active seasons (65% of locations and 51% of time for bear 08; 75% of locations and 63% of time for bear 25) followed by avalanche chutes and sidehill climax and old-growth forests. Rank testing between quality/quantity indices (food plant nutrient content, biomass, berry abundance) and grizzly bear use indicated that movements were generally correlated with food availability at the higher, or BHU, level of the classification (rs=0.61 and 0.83, p<.05 for bears 08 and 25, respectively). Salmon (Oncorhynchus spp.) and insects were the only common non-plant items in a diet of over 3 0 species, although food habits differed between bears. It is concluded that, although food plays a critical role in habitat selection of coastal grizzly bears, a relatively rich environment precludes the need for individuals to forage optimally at a micro-habitat level. Except for the early spring and late fall, food can be found in a number of units that collectively meet life requisites. These analyses were used in combination with other use and habitat quality information to develop seasonal habitat values. Assigned values allowed predictions about the effects of forest management practices on habitat capability. / Forestry, Faculty of / Graduate