One hundred and ten children aged 6 months to 11 years were randomly treated with halofantrine (HF), sulfadoxine-pyrimethamine (S-P) or chloroquine (CQ) for acute uncomplicated Plasmodium falciparum malaria in an endemic area of south-western Nigeria. The response of infection to treatment in each child was monitored for 14 d. The mean fever clearance times were 1.9 d (n = 36), 1.6 d (n = 27), and 1-7 d (n = 28) for children treated with HF, S-P and CQ, respectively. The parasite clearance times were 3.4 d (n = 39), 4.4 d (n = 24) and 4.1 d (n = 15) in the 3 groups of children. The cure rate at day 7 was 92.3% (36/39) in children treated with HF, 72.7% (24/33) in those treated with S-P, and 39.5% (15/38) in those treated with CQ. By day 14, 4 of 36 (11.1%) parasitologically cured patients treated with HF had experienced recrudescences. The corresponding figures among children treated with S-P or CQ were 8.3% and 13.3%, respectively. The 3 drugs were well tolerated. The results of the study showed a further decline in the sensitivity of P. falciparum infections to CQ, while HF and S-P remained relatively effective in the treatment of malaria in south-west Nigeria.