From November 1991 to November 1992 a prospective, descriptive study of malaria epidemiology was conducted in a Karen population on the western border of Thailand. Two study groups were selected at random and more than 80% of the subjects were followed for one year. In Group 1, comprising 249 schoolchildren (aged 4-15 years), daily surveillance for illness was combined with fortnightly malaria surveys. These children experienced 1.5 parasitaemic infections per child-year (95% confidence interval [CI] 1.3-1.7), of which 68% (193/285) were symptomatic (Plasmodium falciparum 84%, P. vivax 57%). The estimated pyrogenic densities were 1460/microL for P. falciparum and 181/microL for P. vivax. In Group 2, comprising subjects of all age from 428 households, malaria was diagnosed during two-monthly surveys, at weekly home visits, and otherwise by passive case detection. Malaria and splenomegaly prevalence rates were low in all age groups (spleen index 2-9%; P. falciparum prevalence rate 1-4%; P. vivax 1-6%). Group 2 subjects had 1.0 infections per person-year (95% CI 0.9-1.1), most of which were symptomatic (312/357; 87%). Malaria infections clustered in households. Overall, P. vivax caused 53% and P. falciparum 37% of the infections (10% were mixed), but whereas P. vivax was most common in young children, with a decline in incidence with increasing age, P. falciparum incidence rates rose with age to a peak incidence between 20 and 29 years, although the risk of developing a severe malaria decreased with increasing age. There was no death from malaria during the study. P. falciparum infections were more common in males, subjects with a history of malaria before the study, and in those who had travelled outside their village. These findings suggest a higher transmission rate for P. vivax than P. falciparum, although adults still suffered symptomatic malaria due to both species. The 2 malaria parasites found in this area contribute approximately 50% of infections each, but their clinical epidemiology is very different.