High-throughput screening (HTS) is changing as more compounds and better assay techniques become available. HTS is also generating a large amount of data. There is a need to rationalize the HTS process, because, in some cases, the screening of all available compounds is not economically feasible. In addition to the selection of promising compounds, there is a need to learn from the data that we collect. In this paper, we use a data-mining method, recursive partitioning, to help uncover and understand structure-activity relations and to help biology and chemistry experts make better decisions on which compounds to screen next and better characterize. The sequential-screening process is presented and the results of applying that process to 14 G-protein-coupled receptor assays are reported.