Invasion of erythrocytes by Plasmodium merozoites is mediated by a series of specific receptor-ligand interactions. The interaction of the Duffy binding protein (DBP) with its receptor, the Duffy blood group antigen, is essential for P. vivax invasion of human erythrocytes. Due to its critical role in maintaining blood-stage P. vivax infection, DBP is a candidate for inclusion in a vaccine against P. vivax. Naturally occurring antibodies to DBP are prevalent in individuals living in areas of high malaria endemicity, but individuals show distinct quantitative and qualitative differences in their serological response to the antigen. Polymorphisms in the dbp gene are clustered in its ligand domain, suggesting immune selection pressure. Previous studies using pooled sera from residents of Papua New Guinea showed a correlation between anti-DBP antibody titer and inhibition of erythrocyte binding by DBP. Serum samples were collected from individuals residing in an area of PNG highly endemic for P. vivax. Sera were tested for anti-DBP titer by ELISA and for ability to block binding of DBPII to erythrocytes by means of an in vitro assay. Among high responders to the DBP, a range of inhibition from complete to none was observed indicating that some anti-DBPII antibodies are considerably more effective than others at blocking binding of the ligand domain. Pooling the inhibitory serum samples did not enhance inhibition suggesting that relatively few neutralizing epitopes are present on the DBP ligand domain. Polymorphisms in DBP cluster in the binding domain of region II, which suggests that they function as a mechanism of immune evasion. To determine if these polymorphisms affect antibody recognition of DBPII, we tested whether antisera to rDBP (Sal1 variant) had comparable inhibition of binding activity of two different DBPII variants from PNG. Antisera recognized the variant DBPII types in a strain-specific manner. Differences in antigenic character were mediated by single variant residues, with multiple polymorphisms having an additive effect. These results suggest that polymorphism affects the antigenic character of DBPII and may contribute to the strain-specific immunity to P. vivax observed in endemic populations. To investigate the roles of conserved and polymorphic residues in region II, I used site-directed mutational analysis to identify residues critical to receptor recognition. Alanine substitutions were created in 75 surface-predicted residues in the central portion of region II. Effects of substitution ranged from no change to complete abrogation of the erythrocyte binding activity for the ligand domain. Mutations that completely abrogated erythrocyte binding were mostly distributed in discontinuous clusters with mutations in flanking residues conferring partial loss of binding function. Alanine substitution of polymorphic residues from the hypervariable segment of region II generally did not impair binding function. These data will help identify epitopes recognized by inhibitory antibodies and characterize conserved, critical binding determinants potentially useful for vaccine design.