OBJECTIVE: Parasites may recur asymptomatically after initial clearance by antimalarial treatment. Current guidelines recommend treatment only when patients develop symptoms or at the end of follow-up. We wanted to assess prospectively the probability of becoming symptomatic and the risks of this practice. METHODS: We analysed data collected in 13 trials of uncomplicated paediatric malaria conducted in eight sub-Saharan African countries. These studies followed all cases of post-treatment asymptomatic parasitaemia until they developed symptoms or to the end of the 28-day follow-up period, at which time parasite genotypes were compared to pre-treatment isolates to distinguish between recrudescences and new infections. RESULTS: There were 425 asymptomatic recurrences after 2576 treatments with either chloroquine, sulfadoxine/pyrimethamine or amodiaquine, of which 225 occurred by day 14 and 200 between day 15 and day 28. By day 28, 42% developed fever (median time to fever = 5 days) and 30% remained parasitaemic but afebrile, while 23% cleared their parasites (outcome unknown in 4%). Young age, parasitaemia >/=500 parasites/microl; onset of parasitaemia after day 14, and treatment with amodiaquine were the main variables associated with higher risk of developing fever. CONCLUSION: In areas of moderate to intense transmission, asymptomatic recurrences of malaria after treatment carry a substantial risk of becoming ill within a few days and should be treated as discovered. Young children are at higher risk. The higher risk carried by cases occurring in the second half of follow-up may be explained by falling residual drug levels.