Activity time budgets in apex predators have been proposed as indicators of population status relative to resource limitation or carrying capacity. We used archival time-depth recorders implanted in 15 adult female and 4 male sea otters (Enhydra lutris) from the northernmost population of the species, Prince William Sound, Alaska, USA, to examine temporal patterns in their foraging behavior. Sea otters that we sampled spent less time foraging during summer (females 8.8 hr/day, males 7.9 hr/day) than other seasons (females 10.1–10.5 hr/day, males 9.2–9.5 hr/day). Both sexes showed strong preferences for diurnal foraging and adjusted their foraging effort in response to the amount of available daylight. One exception to this diurnal foraging mode occurred after females gave birth. For approximately 3 weeks post-partum, females switched to nocturnal foraging, possibly in an effort to reduce the risk of predation by eagles on newborn pups. We used multilevel mixed regression models to assess the contribution of several biological and environmental covariates to variation in the daily foraging effort of parous females. In the random effects only model, 87% of the total variation in foraging effort was within-otter variation. The relatively small among-otter variance component (13%) indicates substantial consistency in the foraging effort of sea otters in this northern population. In the top 3 models, 17% of the within-otter variation was explained by reproductive stage, day length, wind speed, air temperature and a wind speed × air temperature interaction. This study demonstrates the potential importance of environmental and reproductive effects when using activity budgets to assess population status relative to carrying capacity.