OBJECTIVES: Previous studies of a fixed combination including cotrimoxazole, rifampicin, and isoniazid (Cotrifazid) showed efficacy against resistant strains of Plasmodium falciparum in animal models and in small-scale human studies. We conducted a multicentric noninferiority trial to assess the safety and efficacy of Cotrifazid against drug-resistant malaria in Papua New Guinea. DESIGN: The trial design was open-label, block-randomised, comparative, and multicentric. SETTING: The trial was conducted in four primary care health facilities, two in urban and two in rural areas of Madang and East Sepik Province, Papua New Guinea. PARTICIPANTS: Patients of all ages with recurrent uncomplicated malaria were included. INTERVENTIONS: Patients were randomly assigned to receive Cotrifazid, mefloquine, or the standard treatment of quinine with sulfadoxine-pyrimethamine (SP). OUTCOME MEASURES: Incidence of clinical and laboratory adverse events and rate of clinical and/or parasitological failure at day 14 were recorded. RESULTS: The safety analysis population included 123 patients assigned to Cotrifazid, 123 to mefloquine, and 123 to quinine + SP. The Cotrifazid group experienced lower overall incidence of adverse events than the other groups. Among the efficacy analysis population (72 Cotrifazid, 71 mefloquine, and 75 quinine + SP), clinical failure rate (symptoms and parasite load) on day 14 was equivalent for the three groups (0% for Cotrifazid and mefloquine; 1% for quinine + SP), but parasitological failure rate (P. falciparum asexual blood-stage) was higher for Cotrifazid than for mefloquine or quinine + SP (9% [PCR corrected 8%] versus 0% and 3%, respectively [p = 0.02]). CONCLUSION: Despite what appears to be short-term clinical equivalence, the notable parasitological failure at day 14 in both P. falciparum and P. vivax makes Cotrifazid in its current formulation and regimen a poor alternative combination therapy for malaria.