There is a pressing need to quantify the risks of renewable energy developments such as offshore wind farms for protected populations. However, assessments are often based on incomplete data, or fail to consider variation in risk between sexes and at different times of year. We tracked northern gannets foraging from the world's largest colony (Bass Rock, Scotland) across five consecutive breeding seasons. We examine how seasonal and sex differences in behaviour affect the collision risk from planned and operational wind farms within their foraging range and assess the likely consequences for long-term population viability. Both sexes made shorter trips during chick-rearing than prior to chick-hatching, spent a greater proportion of time within wind farm sites and had an eight times greater potential collision risk during chick-rearing. Females made longer trips than males at both these times of year, flew higher and spent more time within wind farm sites, leading to three times greater collision risk for females. After accounting for the potential additional mortality from collisions, and assuming that the death of a parent also led to the loss of its offspring, the breeding population was projected to increase by 3.57% (95% CI: 2.16–5.15%) per year, compared with 6.56% (95% CI: 4.59–8.73%) in the absence of turbines, suggesting a negligible effect on population viability. However, additional mortality could result in greater immigration from neighbouring colonies, potentially affecting their viability and highlighting a need for research within a metapopulation framework to assess the impacts of offshore wind developments on vulnerable species across multiple connected sites.