The effects of Schistosoma japonicum infection on the concentrations of zinc in serum, liver, spleen and muscle and on the concentrations of retinol in serum and liver were studied in 48 pigs. Twenty-four of the pigs were each infected by intramuscular inoculation with 2000 cercariae of S. japonicum in medium and the rest were similarly inoculated with parasite-free medium, as controls. On each of weeks 4, 11, 17 and 24 post-inoculation (PI), 12 pigs (six of which were infected) were killed. Tissue samples were collected at necropsy. Blood samples were taken prior to infection and at necropsy from all pigs, and bi-weekly from the pigs killed 24 weeks post-infection. In an analysis of variance in which serum retinol was the dependent variable, the interaction infection x time was found to be significant (P = 0.009). The main reason for this significance was that the concentration of retinol in the sera collected from infected pigs at necropsy at 11 weeks PI was significantly lower than in the control pigs killed at the same time (P = 0.02). Although, overall, infection led to higher zinc concentrations in the liver (P = 0.04) and spleen tissue (P = 0.01), it had no apparent effect on liver retinol, muscle zinc or serum zinc. However, among the pigs which were tested bi-weekly, serum zinc was consistently lower in the infected pigs than in the controls (P = 0.01). The transient declines seen in the concentrations of retinol and zinc in sera from the infected pigs were not accompanied by similar changes in the tissue concentrations, and may reflect an acute-phase response to infection. Schistosoma japonicum infection in pigs is considered a useful model of S. japonicum infection (and probably also of S. mansoni infection) in humans. Similar effects, if they occur in the human infections, may lead to misclassification of vitamin-A and zinc status in endemic populations if this status is based on serum retinol and serum zinc.