This article argues that the use of genome data and high-throughput systems for screening compounds for antimicrobial activity has not been as successful as hoped, that screening compounds against whole parasites (even if the mechanism of action is not known) has been more effective at identifying potential antimalarials, and that whole-parasite-based screening is more likely to identify compounds that act on more than one target (making it harder for the parasites to develop resistance), but that searching for molecular targets still has a very useful role. It does so in the light of a research paper in the same issue of the journal (M. Rottmann et al.: Spiroindolones, a Potent Compound Class for the Treatment of Malaria. Science (2010) 329 (5996) 1175-1180. DOI:10.1126/science.1193225) which describes the rapid development of a preclinical antimalarial candidate by parasite-based screening.