SPf66, a synthetic peptide Plasmodium falciparum vaccine, did not protect young Gambian children against clinical attacks of malaria. Nevertheless, Gambian children who had been vaccinated with SPf66 and who were parasitaemic at the end of the first malaria transmission season after vaccination had significantly fewer detectable P. falciparum genotypes than control children, as determined by polymerase chain reaction analysis of 3 polymorphic loci--the msp1 block 2 repeat region, the msp2 repeat region, and the R11 region of the glutamate-rich protein gene (glurp). Geometric mean numbers of genotypes were 1.66 vs. 1.87, 1.95 vs. 2.43, and 1.21 vs. 1.50 for msp1, msp2 and glurp, respectively (P = 0.31, P = 0.04 and P < 0.01). Differences between groups became a little more marked for msp1 and msp2 when children with symptomatic malaria were excluded. No significant difference was found between parasites obtained from SPf66-vaccinated or control children in the prevalences of amino acid alleles at positions 44 and 47 in the 11 amino acid sequence of the merozoite surface protein 1 molecule, which is present in SPf66. The reduction in the number of genotypes observed could not be explained by a difference in parasite densities between SPf66-vaccinated and control children, as geometric mean parasite densities were almost identical in the 2 groups. These observations suggest that SPf66 vaccine may have induced an immune response which reduced the incidence of new infections in immunized children or accelerated the rate of clearance of parasites of individual genotypes. However, no reduction in the prevalence or density of parasitaemia was recorded in SPf66-vaccinated children, suggesting the existence of some kind of density-dependent mechanism for controlling low levels of malaria parasitaemia.