The interaction between malnutrition and malaria is complex and there is evidence that malnutrition decreases the susceptibility to malaria. To investigate the relation between anthropometric measurements and subsequent malaria morbidity and to examine whether the effect observed was due to interaction with host immunity, we followed for 1 y a cohort of 136 children aged 10 to < 120 mo in Wosera, East Sepik Province, Papua New Guinea. At baseline, 21% were stunted, 10% were wasted, and 5% were both stunted and wasted. After adjustment for age and use of bed nets, height-for-age z score (HAZ) at baseline predicted the number of clinical episodes of falciparum malaria during the following year: incidence rate increased with increasing HAZ. Humoral responses to specific malarial antigens were lowest in the wasted children. The prevalence of lymphoproliferative responders was not significantly different between well-nourished and undernourished children. In contrast, the prevalence of cytokine producers was higher in the undernourished than in the well-nourished children. Our findings support the view that stunting but not wasting protects against falciparum malaria. The mechanism may be related to an improved ability of malnourished children to produce certain cytokines in response to stimulation by specific malarial antigens.