Recent population genetic studies on the malaria parasite Plasmodium falciparum have confirmed that selfing is more frequent where the transmission rate is lower, with inbreeding coefficients estimated to be 0.33 and 0.92 for sites in Tanzania and Papua New Guinea (PNG), respectively. These geographical differences in Plasmodium mating patterns have been linked to the rate of spread of chloroquine resistance CQR, which, according to some measures, has been slower in Tanzania than in PNG. It has been proposed that the former observation explains the latter, although the theoretical argument linking the two is based on limited simulation studies. Taking a more analytical approach here, we first establish the relevant relationship between the coefficient of inbreeding (F, within loci) and the recombination rate r, between loci, defining an 'effective recombination rate', r = r (1-F. We then show that the emergence of multigenic drug resistance can indeed be slowed (or even quickened) by more outcrossing, but only when resistance is determined by two or more genes, none of which independently confers significant protection. The resistance genes should both be initially rare, and subject to low selection pressure. The analysis does not completely discount the hypothesis that inbreeding significantly influences the spread of CQR, but we show that it can only do so under a restrictive set of conditions, and that these conditions are not satisfied by some laboratory and field data. We discuss some of the wider implications of these results for the evolution of multigenic resistance.