SummaryFlying birds have been documented to respond in different ways to the presence of wind farms. Such responses are species‐ and site‐specific, with wind farm design playing an important role. Between 2009 and 2014, the length of rows of wind turbines within our study area increased from 3·4 km to ∼7·5 km, and the total area occupied by wind farms increased from 6·47 km2to 14·21 km2. This area is located on an important migratory corridor in southern Mexico.We used marine radar and hawk‐watch monitoring stations to collect data during six consecutive autumn seasons from a single wind farm. We analysed the response of migrating raptors to the presence of two new wind farms by comparing the mean bearing of flight trajectories and the number of intersections km−1of trajectories with wind farm areas between 2009–2011 and 2012–2014, representing the pre‐ and post‐construction stages of the new wind farms.Mean raptor count was >600 000 individuals for the six seasons. The most abundant species were Turkey VultureCathartes aura, Swainson's HawkButeo swainsoniand Broad‐winged hawkButeo platypterus. Between 79% and 97% of migration occurred in October. Radar monitoring overlapped with the peak migratory activity each season.We observed significant differences between periods, involving more scattering in flight bearings and less intersections km−1of trajectory in the post‐ than in the pre‐construction period, implying an avoidance of the new wind farms.Synthesis and applications. We show that migrating raptors adjusted their flight trajectories to avoid new wind farms, but also discuss the extent and limitations of our findings. Our results from our hawk‐watch monitoring station, which represent the first published account about the seasonality and intensity of raptor migration in the area, could be used by decision‐makers for careful planning of future wind energy developments in the area. Our results might aid in the conservation of those species of raptors that migrate through the Isthmus of Tehuantepec.