In an area of Papua New Guinea with high prevalence of Plasmodium falciparum (39.6%), Plasmodium vivax (18.3%), and Plasmodium malariae (13.8%), cross-sectional analysis found P. falciparum infection to be independent of the other species despite heterogeneities in transmission. Plasmodium vivax and P. malariae infections were negatively correlated. Plasmodium malariae infection was positively associated with homologous infection four months previously and with prior P. falciparum, but not P. vivax infection. There were no other indications that any Plasmodium species protected against heterologous infection. Prospective analysis of health-center morbidity supported the idea that P. malariae infection protects against disease, but indicated greater protection against non-malaria than P. falciparum-associated fevers. Plasmodium vivax appeared to protect against P. falciparum disease but not against other forms of morbidity. Covariate adjustment had considerable effects on estimated relationships between species, and confounding variables may account for many differences among reports of inter-species interactions in human malaria.