OBJECTIVE: To compare the parasitological failure rates of under-fives and pregnant women with parasitaemia treated with chloroquine (CQ) or sulphadoxine-pyrimethamine (SP). METHODS: During a clinical trial of CQ, SP, amodiaquine (AQ) and SP plus AQ combination for malaria treatment in pregnant women in Ghana, a parallel study of treatment of children below 5 years of age with symptomatic malaria with CQ and SP was undertaken. Four hundred and fifty pregnant women with malaria parasitaemia and 203 children with malaria parasitaemia were randomized to receive CQ or SP. They were followed up and parasitological failure by days 14 and 28 after the start of treatment was assessed. RESULTS: Polymerase chain reaction (PCR)-uncorrected parasitological failure rates by day 28 after the start of treatment with CQ were 58.5% (55/94), 38.5% (45/117), 31% (13/42) and 8.2% (4/49) in children, primigravidae, secundigravidae and multigravidae, respectively. For those treated with SP the rates by day 28 were 36.4% (32/88), 27.1% (29/107), 6.1% (3/49) and 3.8% (2/52) in children, primigravidae, secundigravidae and multigravidae, respectively. In both CQ and SP treatment arms, children were twice as likely to experience recrudescence as pregnant women (RR 2.1 [95% CI 1.6-2.6] P < 0.0001) by day 28 after the start of treatment. CONCLUSIONS: Parasitological failure rates were significantly lower in asymptomatic pregnant women, particularly in multigravidae, compared with symptomatic children. Reliance on drug sensitivity results observed in children only to decide on antimalarial regimes for pregnant women may not be appropriate.