Artemether-lumefantrine is being promoted by the World Health Organization for treating uncomplicated malaria. It is expensive. We sought evidence of its superiority over existing treatment regimens. To compare artemether-lumefantrine with other antimalarial drugs for treating uncomplicated falciparum malaria. We searched the Cochrane Infectious Diseases Group specialized trials register (April 2002), the Cochrane Controlled Trials Register (Issue 2, 2002), MEDLINE (1966 to April 2002), EMBASE (1988 to April 2002), conference proceedings, and reference lists of articles. We contacted experts in malaria research and the pharmaceutical company that manufactures artemether-lumefantrine. Randomized and quasi-randomized trials comparing artemether-lumefantrine administered orally with standard treatment regimens (single drug or combination). Two reviewers independently applied inclusion criteria to potentially relevant trials, assessed trial quality, and extracted data. Parasitaemia on day 28 (day 42 for sulfadoxine-pyrimethamine and day 63 for mefloquine) was the primary outcome. Adverse event information was collected from the studies. Eight trials (2117 participants) met the inclusion criteria. In the four studies against single agents, failure rates for artemether-lumefantrine tended to be higher in comparisons against sulfadoxine-pyrimethamine, halofantrine, and mefloquine. This difference was statistically significant for mefloquine. When compared with chloroquine, artemether-lumefantrine was better in two studies, but the failure rate for chloroquine at these sites was over 50%. All single agent studies used four doses of artemether-lumefantrine. In comparisons against combination treatment, three trials tested artemether-lumefantrine against mefloquine-artesunate and showed that artemether-lumefantrine was inferior for day 28 cure (Relative Risk 6.33, 95% confidence interval 3.08 to 13.01). If this comparison is confined to the two trials where participants received six doses, artemether-lumefantrine was associated with higher cure rates, but this was not statistically significant (Relative Risk 4.20, 95% confidence interval 0.55 to 31.93). Artemether-lumefantrine is more effective than chloroquine in chloroquine resistant areas. Artemether-lumefantrine is less effective than mefloquine or mefloquine combined with artesunate. We found no evidence to confirm or refute whether artemether-lumefantrine was better than sulfadoxine-pyrimethamine.