The mechanisms of protection and pathogenesis in malaria are not well understood. Antibodies to sporozoites and asexual erythrocytic stages of the parasite antigens have been shown to have a protective role in experimental model. In human P. falciparum infections IgC has been shown to limit parasitemia. Cytokines may have a protective role in malaria, and may also be involved in the development of clinical disease. The overproduction of tumor necrosis factor (TNF-$\alpha$) induced by soluble malarial antigens (exoantigens) has been implicated in immunopathology. The purpose of the study was two-fold: (1) to assess the immune reactivity of exposed individuals to soluble P. falciparum blood-stage antigens (Pf70) and to determine if the presence of Pf70 antibody is associated with protection or increased risk of disease; and (2) to further investigate the relationship between TNF-$\alpha$ and disease. The study was carried out on (361) subjects in Kinshasa (Zaire). The study was designed as cross-sectional and was both community and hospital based. Total anti-asexual P. falciparum blood stage antibodies, antibodies to Pf70 antigen, and circulating Pf70 antigen levels were measured by immunofluorescence and by ELISA in adults and children from the community. The same parameters and TNF-$\alpha$ were measured in children admitted to the hospital with different degrees of malaria severity and children with other acute infections, and were compared to children with mild malaria, asymptomatic carriers and healthy children from the community. The pattern of malarial infection in the community was also determined. Ninety percent of the subjects had antibodies to Pf70 antigen. Antibody levels against Pf70 were higher among children with low parasitemias (P $<$ 0.05), suggesting a protective role for anti-Pf70. However both anti-Pf70 and IFA anti-asexual blood-stages antibodies failed to protect against severe malaria and clinical disease. TNF-$\alpha$ concentrations were lower in asymptomatic individuals as compared to symptomatic subjects (P $<$ 0.05), although the cytokine levels did not correlate with severity in acute malaria. Pf70 antigen was detected in 91% of parasitologically-confirmed infections and 71% of microscopically-negative subjects. The high apparent prevalence of Pf70 antigen among blood-smear negative individuals indicates either low level infection not detected by microscopy or cross reactivity of the rabbit polyclonal antibody used to detect Pf70 with other plasma components