A novel strategy for isolating potential anti-tumor drugs is presented. It is predicated on the idea that future anti-tumor drugs will be specific inhibitors of the signal-transduction pathways responsible for cell proliferation. Briefly, retroviral vectors are used to introduce focus-forming oncogenes into a test population of target cells, which are grown to confluence and treated with signal-transduction inhibitors. The inhibitors are screened for the ability to suppress the development of transformed foci without killing the confluent monolayer of non-transformed quiescent cells. For this work, a panel of inhibitors was first screened against the oncogene ras. The protein kinase C (PKC) inhibitor CGP 41251 and the protein tyrosine kinase (PTK) inhibitor CGP 45047 suppressed ras-induced focus formation and left a viable monolayer of quiescent cells. Focus inhibition was reversible; conversely, drug addition to developing foci retarded further expansion. CGP 41251 generally blocked proliferation of ras or control cells, suggesting that oncogenes cannot substitute for PKC. PTK inhibitors erbstatin and CGP 520 and phosphatase inhibitor okadaic acid failed to inhibit focus formation at concentrations toxic to the monolayer. Lavendustin A and CGP 47778A showed neither focus inhibition nor toxicity. In the complementary screen, a single inhibitor (CGP 41251) was tested against several oncogenes, including src, raf and polyomavirus middle T antigen. Focus formation by all oncogenes was suppressed. The strategy has several advantages over current drug-screening assays, and it can be adapted to large-scale screening with many drugs and many oncogenes.