The rationale for exchange blood transfusion (ET) in severe falciparum malaria is threefold: reduction of parasitaemia, reduction of presumptive 'toxic' factors, and improvement of the rheological quality of the blood. We evaluated the records of 61 patients treated with ET to describe the present status of malaria treatment in Germany, Austria and Switzerland and to assess the efficacy of ET. Clinical data of 61 patients treated with ET were compared to data of 63 patients treated in 2 hospitals where ETs were generally not performed. We found that exchange transfusion is applied according to the clinician's subjective impression rather than strict guidelines. Logistic regression analysis adjusting for the differences in clinical parameters between patients treated with or without ET did not identify treatment as a prognostic indicator (odds ratio for relative risk of death with ET: 1.3; 95% CI: 0.4-4.9). Exchange transfusion did not significantly improve the unfavourable prognosis in cases of severe falciparum malaria. However, failure to reach statistical significance may be due to the retrospective design of the study and therefore non-systematic approach.