In the past, assessment of drug efficacy against Plasmodium falciparum malaria has been performed by microscopy, screening for parasites in blood smears. However, in areas of high endemicity, reappearing parasites might be derived from new inoculations and could be classified falsely as treatment failures. Recently, a number of studies have used polymerase chain reaction (PCR) analysis of detectable parasites after drug administration to discriminate new infections from true recrudescence. The feasibility, high sensitivity and high resolution of this technique proves that it will be practical and highly valuable in studies on both drug resistance and vaccine efficacy as well as the testing of novel antimalarial drugs. In this article, Georges Snounou and Hans-Peter Beck discuss the uncertainties in the interpretation of data inherent to the technical limitations of the PCR technique, and the constraints imposed by the biology of the parasite. They suggest that although genotyping can provide strong evidence for differentiating between true recrudescence and reinfection, it must be interpreted with caution. They also propose strategies that might help minimize these uncertainties.