Migration is a fundamental aspect of the ecology and evolutionary history of many animals, driven by seasonal changes in resource availability and habitat structure. Seabird migration has been investigated extensively in highly seasonal temperate and polar environments. By contrast, the relationships between migration and seasonal environmental changes have rarely been studied in tropical marine habitats. The sooty tern Onychoprion fuscatus is the most abundant tropical seabirds, and has been ranked as the most important tropical species in terms of its annual estimated consumption of marine resources. We used global location sensing (GLS loggers) to describe for the first time the year-round at-sea distribution and activity patterns of sooty terns from a large breeding colony in the western Indian Ocean (Bird Island, Seychelles). While breeding, they foraged within 1,074 ± 274 km of the colony. After breeding, birds undertook an extensive post-breeding migration throughout the Indian Ocean; average distances traveled exceeded 50,000 km per individual. Sooty terns used mainly four distinct core oceanic areas during the non-breeding period; in the Bay of Bengal (A), northeast to an area straddling the Chagos-Laccadive plateau (B), southeast to an area on each side of the 90 East Ridge (C) and southwest to an area around Comoros (D). Individuals exhibited a high degree of fidelity to these core areas in successive years. We also established that they performed an unusual behavior for a non-Procellariiformes seabird; most individuals undertook a 1-month pre-laying exodus, during which they foraged in a specific area c. 2,000 km to the south-east of the colony. Year-round at-sea activity of sooty terns revealed that they spent only 3.72% of their time in contact with seawater, so indicating that they must sleep in flight. Activity parameters exhibited seasonal (breeding vs. non-breeding periods) and daily variations; they notably never land on the water at night. In the Seychelles, breeding sooty terns are threatened by commercial egg harvesting. Our discovery of extremely wide non-breeding at-sea distribution highlights the risk of other threats during their non-breeding period, such as over-fishing, marine pollution and climate change.