Identifying areas with a high risk of collision with wind turbines is crucial for the conservation of large soaring birds. However, studies often rely on different wind turbine data sources, and the effect of this heterogeneity on collision risk estimation remains poorly understood. To address this gap, we combined GPS and accelerometer data from six Griffon Vultures Gyps fulvus tracked in Sardinia (Italy) with three wind turbine data sets differing in accuracy and completeness. We estimated the vultures’ foraging grounds and calculated collision risk using each data set, including one based on aerial imagery (unbiased), one from OpenStreetMap (OSM), and a third from published literature. Results showed that turbines mapped from aerial imagery overlapped with 18.7% of foraging areas, compared with 8.7% using OSM and 15.9% using the third data set. Projections including planned wind farms indicated that 31.4% of current foraging grounds will become at risk. These findings demonstrate that wind turbine data sources significantly influence estimates of collision risk. The need for reliable, accessible, and regularly updated turbine maps to support effective conservation planning, guide mitigation actions such as selective shutdowns, and monitor cumulative impacts over time are highlighted.