The ambulatory management of imported Plasmodium falciparum malaria is controversial because criteria for safe selection of patients are imprecise. The aim of the present study was to investigate the evolution and outcome of patients diagnosed with Plasmodium falciparum malaria at a Belgian referral institute in order to assess the safety of the institute's current selective ambulatory management protocol. From 2000 to 2005, all patients diagnosed with P. falciparum infection at the Institute of Tropical Medicine and the University Hospital of Antwerp were enrolled prospectively. Ambulatory treatment was offered to nonvomiting patients if they exhibited none of the 2000 World Health Organization criteria of severity and had parasitemia below 1% at the initial assessment. The treatment of choice was quinine (plus doxycycline or clindamycin) for inpatients and atovaquone-proguanil for outpatients. P. falciparum malaria was diagnosed in 387 patients, of whom 246 (64%) were Western travelers or expatriates and 117 (30%) were already on antimalarial therapy. At diagnosis, 60 (15%) patients had severe malaria. Vital organ dysfunction was initially seen in 34 and developed later in five others. Five patients died. Of the 327 patients initially assessed as having uncomplicated malaria, 113 (35%) were admitted immediately; of these, 4 developed parasitemia >/=5% at a later stage but without any clinical consequence. None of the 214 individuals initially treated as outpatients experienced any malaria-related complications, including 10 who were admitted later. Vital organ dysfunction was observed in only 2 of the 214 patients with initial parasitemia <1% who had not taken antimalarial agents (both patients had impaired consciousness at presentation). Ambulatory treatment is safe in treatment-naive malaria patients with parasitemia <1% who do not vomit and who do not exhibit any criteria of severe malaria.