This study was conducted in 2003 as part of the training of laboratory technicians in the use of rapid diagnostic tests (RDTs) for malaria and to evaluate these tests in Madagascar in field conditions for the first time. Two types of RDT were used separately. The dipstick (Optimal-I) that detects circulating pLDH was tested in 168 patients with clinically suspected malaria (fever or recent history of fever) at primary health centers. Microscopy confirmed malaria in 93/168 (55.4%) cases. Monoparasitic P. falciparum infection was identified in 86/93, P. malariae in 3/93, P. vivax in 3/93 and P. ovale in 1/93. A positive Optimal-I test was a highly sensitive indicator of P. falciparum infection with parasitemia exceeding 500 trophozoites/mul (sensitivity of 97.2%; with a specificity of 100%); it also confirmed 6/7 cases of non-P. falciparum malaria. A community malaria survey used the Malaria Hexagon dipstick (detecting P. falciparum-specific HRP2) for 273 patients: 17 (6.2%) RDT tests were positive, and 16 (5.9%) microscopic tests. Although this dipstick did not detect the only case of infection with P. vivax, its specificity was 100% for detection of P. falciparum infection. Installing microscopes and qualified microscopists in the health centers of the one hundred and eleven districts in Madagascar would be extremely difficult, but our results show that RDT is an effective alternative diagnostic tool for daily use as well as for sporadic malaria epidemics. The revised antimalarial treatment policy, involving a drug ten to twenty times more expensive than chloroquine, demonstrates the need to improve malaria diagnosis: presumptive treatment has become prohibitively expensive. RDT can be used to improve malaria case management at the primary heath centers in Madagascar. We discuss the choice of RDTs.