Many three-dimensional (3D) virtual screening concepts, like automated docking or pharmacophore searching, rely on the calculation of a "bioactive" or "receptor-relevant" conformation of a molecule to assess its biological activity. We investigated the dependence of the presence of conformations near the "bioactive" conformation on three-dimensional similarity searching with pharmacophore-based correlation vectors (CATS3D approach). Cocrystal structures of 11 target classes served as queries for virtual screening of a database of annotated ligands. Different numbers of conformations were calculated. Single 3D structures were obtained using the 3D structure generator CORINA and conformational ensembles by the conformation generation program ROTATE. This approach was able to reproduce conformations for high resolution cocrystal structures. For virtual screening we found that using only the CORINA-generated single conformation already resulted in a significant enrichment of isofunctional molecules having the same biological property profile. This observation was also made for ligand classes with many rotatable bonds. Although more similar conformations were considered to be more similar in the CATS3D description, the impact of using multiple conformations on the enrichment of actives was not as high as expected. CATS3D provides an alignment-free three-dimensional virtual screening approach that is less dependent on the presence of conformations which are close to the "bioactive" conformation of a molecule compared to methods that rely on an explicit three-dimensional alignment of molecules.