The evolutionary selection of malaria parasites within individual hosts is an important factor in the emergence of drug resistance but is still not well understood. We have examined the selection process for drug resistance in the mouse malaria agent Plasmodium berghei and compared the dynamics of the selection for atovaquone and pyrimethamine. Resistance to these drugs has been shown to be associated with genetic lesions in the dihydrofolate reductase gene in the case of pyrimethamine and in the mitochondrial cytochrome b gene for atovaquone. A mouse malaria model for the selection of drug resistance, based on repeated incomplete treatment (RICT) with a therapeutic dose of antimalarial drugs, was established. The number of treatment cycles for the development of stable resistance to atovaquone (2.47 ؎ 0.70; n ؍ 19) was found to be significantly lower than for pyrimethamine (5.44 ؎ 1.46; n ؍ 16; P < 0.0001), even when the parental P. berghei Leiden strain was cloned prior to the resistance selection. Similar results were obtained with P. berghei Edinburgh. Mutational changes underlying the resistance were identified to be S110N in dihydrofolate reductase for pyrimethamine and Y268N, Y268C, Y268S, L271V-K272R, and G280D in cytochrome b for atovaquone. These results are consistent with the rate of mitochondrial DNA mutation being higher than that in the nucleus and suggest that mutation leading to pyrimethamine resistance is not a rare event.D rug-resistant parasites have become a major challenge to malaria control today. The emergence of resistance is the outcome of two related processes: the genetic event that produces resistant mutants within individual hosts and the spread of resistance in populations. While, as the initial event in the emergence of resistance, the evolutionary selection of malaria parasites within an individual host is critical, it is still not well understood.The study of within-host selection of drug resistance benefits from animal models of malaria infection, as it allows genetic and physiological manipulations in vivo. Mutations can be selected without mosquito passage (i.e., without meiotic recombination) by exposure of large numbers of malaria parasites to certain drug concentrations. Drug-resistant Plasmodium falciparum organisms have been isolated in in vitro cultures (1-3), but drug-resistant Plasmodium berghei, Plasmodium yoelii, and Plasmodium chaubadi can be isolated in vivo in mice. Two general approaches to study in vivo antimalarial resistance selection have been employed and are compared by Peters (4): the serial technique (ST), in which drug dose is gradually increased after each passage, and the 2% relapse technique (2%RT), in which a single and high drug dose is administered at the time of each passage. While these approaches have proven to be effective in the selection of stable resistant strains, both are unnatural treatment regimens and thus might not be suitable models for the study of within-host emergence of antimalarial drug resistance.More recently, we have intr...