Among seabirds, lesser black-backed gulls (Larus fuscus) are considered to be at high risk of colliding with offshore wind turbines. In this respect, we used GPS tracking data of lesser black-backed gulls caught and tagged in two colonies along the Belgian North Sea coast (Ostend and Zeebrugge) to study spatial patterns in the species’ presence and behaviour in and around the Thornton Bank offshore wind farm (OWF). We found a significant decrease in the number of GPS fixes of flying birds from up to a distance of at least 2000 m towards the middle of the wind farm. Non-flying birds showed a similar avoidance of the wind farm interior, yet presence strongly peaked right at the wind farm’s edge, demonstrated to represent gulls perching on the outer turbine jacket foundations. The findings of this study reveal a strong within-wind farm variability in bird density, a most crucial parameter in collision risk modelling. The method presented here is straightforward and similar studies conducted at other wind farm sites on a range of large gull species (Larus sp.) would allow to assess the potential and species-specific variation in meso-scale response patterns and to gain insight in the underlying ecological incentives, which in turn would provide widely applicable and much-needed input for (cumulative) collision impact assessments.