Habitat models for bears were developed using different spatial resolution data.
We compared vegetation data from CORINE, forest map at 1:50,000 scale and LiDAR.
Resolution affects spatial scale at which bear is estimated to perceive landscape.
Reliability of coarse-resolution models depends on indirect non-vegetation factors.
Model performance and credibility increases with resolution and is best with LiDAR.