The in vivo response of falciparum malaria to oral chloroquine and quinine was evaluated in two identical hospital-based, comparative open trials carried out 2 years apart in the same seasonal period at a hospital located in the highlands of Northern Burundi. Children aged 0-14 with uncomplicated falciparum malaria were administered either chloroquine, at 25 mg/kg over 3 days, or quinine, at 10 mg/kg per 8 hourly for 5 days (alternate allocation) and treatment response was evaluated by the WHO 7-day test. In the first study (1992/1993) 472 patients qualified for analyses (211 in the chloroquine and 261 in the quinine group), as compared to 249 subjects in the second study (1994/1995). In each study, the response to quinine was significantly higher than that to chloroquine (P = 0.004 and < 0.001, respectively). While the response to quinine showed insignificant changes over time (95.8 vs. 92.9%), chloroquine was found to be significantly less effective in the second study as compared to the first (77.8 vs. 63.1%; OR (95% CI) 2.04 (1.21-3.43)). Such decline in chloroquine efficacy was attributable to the age group < 5 years of age, where response to chloroquine decreased from 72.9% in 1992/93 to 56% in 1994/1995. Uncontrolled chloroquine use, which spread after the onset in late 1993 of the still ongoing ethnic fighting, appears to be the most likely reason for such a decrease in chloroquine efficacy. Chloroquine resistance has long been known to be present in the hyperendemic lowlands of Burundi, but no data have so far been reported on the response to antimalarials in the highlands of the country. These findings should be considered when deciding on drug policies for the treatment of falciparum malaria in Burundi.