AimsParenteral artesunate is the preferred first-line treatment for severe malaria. Pre-referral rectal artesunate suppositories are recommended where parenteral treatment is inaccessible. In this study, we compared dihydroartemisinin exposure and model-predicted early parasite clearance following rectal artesunate and intravenous artesunate in African children with severe malaria.MethodsA total of 82 African children with severe malaria participated in a randomized crossover study in Democratic Republic of Congo. Forty children received rectal artesunate (10 mg/kg) while the other 42 received intravenous artesunate (2.4 mg/kg) as the first intervention, and then the other route of administration for the second dose. Blood samples were drawn for drug quantification, and nonlinear mixed-effects modelling was used to evaluate the pharmacokinetic properties of intravenous and rectal artesunate and to simulate expected parasite clearance associated with these routes of administration.ResultsThe mean individually estimated rectal bioavailability of artesunate was 21% but was highly variable (IQR: 8%-35%). Plasma exposure to dihydroartemisinin-the principal and bioactive metabolite of artesunate-did not differ significantly between rectal and intravenous administration. Predicted parasite reduction over the first 12 h was similar for both routes, consistent with previously reported clinical benefits of rectal artesunate.ConclusionsThese findings support the 10-mg/kg dose of rectal artesunate as a pre-referral intervention, which should be more widely deployed in malaria-endemic areas.