Partial migration occurs when only a fraction of a population migrates instead of all individuals. Considered an evolutionary precursor to full migration, understanding why some individuals choose to undertake migration while others do not may serve to inform general migratory theory. While several hypotheses currently exist for explaining the maintenance of partial migration, empirical support for many is limited. To address this gap, we analyzed GPS data acquired from brown pelicans (Pelecanus occidentalis; n = 74), a partially migratory seabird, nesting on six colonies in the South Atlantic Bight over the course of four autumn migrations. We estimated that approximately 74% of pelicans nesting within the study area may be migratory on an annual basis, with the remainder staying within the surrounding marine ecoregion year-round. Mean date of migration initiation was 9 November, although movements occurred from September to December. Results from Cox's proportional hazards modeling indicated significant positive and negative effects of sea surface temperatures and body condition on migration rate, respectively. We suggest that the ontogenetic migration of the primary forage species of brown pelicans from estuarine to pelagic environments causes a seasonal reduction in prey and that pelicans in poor body condition are unable to meet the energetic demands potentially associated with this decrease in prey availability (i.e., the fasting endurance hypothesis of partial migration). Although we did not find evidence for a density-dependent migratory response, the effects of intraspecific competition on migration in pelicans also appear to warrant consideration.