Breeding failure is expected to induce behavioural changes in central place foragers. Indeed, after a failed reproductive attempt, breeding individuals are relieved from having to return to their breeding site for reproductive duties and thus are less constrained than successful breeders in their movements during the remainder of the breeding season. Accordingly, they are expected to adjust their behaviour, travelling longer in distance and/or time to reach foraging grounds. They are also expected to use different foraging areas to decrease local intra-specific competition with successful breeders. We compared the at-sea behaviour and habitat use of successful and failed Indian yellow-nosed albatrosses nesting in Amsterdam Island, Southern Indian Ocean, during 2 chick-rearing seasons. Failed breeders exhibited the same at-sea foraging behaviour, travelling as far and as long as successful breeders. They also spent the same amount of time on their nest between at-sea trips. Nevertheless, habitat models revealed partial spatial segregation of failed breeders, which used specific foraging areas characterized by deeper and colder waters in addition to the areas they shared with successful breeders. Our study shows the importance of combining a range of analytical methods (spatial analysis, behavioural inferences with advanced movement models and habitat models) to infer the at-sea behaviour and habitat use of seabirds. It also stresses the importance of considering individual breeding status when aiming to understand the spatial distribution of individuals, especially when this information may have conservation implications.