The preliminary results of a clinical safety review that partly used the Cochrane methodology are presented. Despite methodological limitations including incomplete databases, this review collated evidence corroborating the benign safety profile of the artemisinin type of compounds. No difference was apparent amongst the various derivatives. At the time the workshop was held, 188 studies had been identified of which 108 (enrolling 9,241 patients) fulfilled criteria for analyses. These included both uncomplicated and severe malaria patients enrolled in either controlled and non-controlled studies as well as healthy volunteers. Safety was assessed by analysing adverse events, as well as clinical laboratory (haematology assessed in 4,062, blood chemistry in 3,893 patients), electrocardiographic (2638 patients) and neurological assessments as reported in the papers. No serious adverse event or severe significant toxicity was reported. Overall, the most commonly reported adverse experiences were gastro-intestinal. Occasional neutropenia (1.3%), reticulocytopenia (0.6%), elevated liver enzymes (0.9%) were reported. Transient bradycardia and prolonged QT interval were reported in circa 1.1% of patients monitored. A neurological assessment was performed primarily in the severe malaria patients. No difference was apparent with respect to quinine. In addition, four cases of neuropsychiatric adverse events were reported in patients receiving concomitant mefloquine.