The development of a method called functional group fingerprints is presented along with large-scale validation studies of a consensus fingerprint searching approach. The validation studies show average rates of retrieval when using fingerprint searches on 538 biological target classes from the MDDR that contain an average of 473 active compounds per class. The consensus method has been shown to improve searching efficiency by recovering 60% less false positives at Tanimoto distances of 0.2 and 0.3, and to retrieve 6.2% more active compounds in the top 1% of the ranked MDDR, as averaged across the 538 target classes used in the study. This demonstrates performance improvements of similarity searches as used in the Hit-to-Lead process.