Individual variation in survival and behavior of American marten (Martes americana) was studied in relation to disease, prey fluctuation, and clear-cutting from 10 January 1986 through 20 August 1987 in Newfoundland, Canada. Thirty-seven of forty marten captured on the study area were telemetered and monitored for part or all of the study. Marten mortality was concentrated in two intervals, fall 1986 and late winter 1987. Mortality during fall 1986 was attributable to encephalitis, while marten deaths during late winter 1987 resulted from predation and starvation attributable to the prey decline. Nonsuppurative encephalitis was first detected 7 October 1986; no further evidence of the disease could be found after 1 November 1986. In early October 1986, declining populations of meadow voles were documented; by June 1987 no voles could be found on the study area. In both mortality periods, young-of-the-year marten had lower survival rates than older marten, and transients survived less well than residents. However, encephalitis appeared to be a less selective mortality agent than the prey decline. Females, considered to be more vulnerable to resource perturbations, had lower survival rates and males higher rates during late winter 1987 than during the disease epizootic. Clear-cutting operations ran from 4 August 1986 through 14 November 1986; 3% (259 ha) of the study area was cut. Marten of all ages avoided clear-cuts during logging operations and for the first nine months afterward. Resident kits made significantly greater use of clear-cuts than older residents and were 3.2 times more likely than older residents to be found within clear-cuts. However, resident kits were 2.6 times and adults 8.3 times more likely to use habitats other than clear-cuts. The decline in prey abundance resulted in several changes in marten movement and spacing behaviors. Intrasexual home range overlap by residents was eliminated. All female residents present before the prey decline either died or abandoned their home ranges. The ratio of transients to resident numbers increased. Recolonization of vacated habitats was slower, and duration of dispersal for females increased. Intruder pressure and mating access appeared to play little role in the observed changes in social spacing. The decline in marten numbers during and after the prey decline appeared to have been partially affected by changes in spacing behaviors.