1. A method is described for the measurement of phenobarbitone (PB) by reversed phase high performance liquid chromatography (h.p.l.c.) from small samples of whole blood dried onto filter paper strips. 2. The disposition of PB given prophylactically to young children with severe malaria on parenteral quinine is contrasted with that in aparasitaemic Kenyan children on no antimalarial drugs. There were no differences in the disposition of PB between the two groups. 3. Peak blood PB concentrations were equal to or greater than 15 mg l-1 in 27% of the patients on quinine and 23% of those not on quinine; a concentration of 10 mg l-1 was achieved or exceeded by 100% and 92% of each group, respectively, and was maintained for 39 +/- 24 h (mean +/- s.d.), and 33 +/- 21 h, respectively. 4. In an open, dose-finding study, the progress of young children with cerebral malaria given prophylactic PB (10 mg kg-1), was contrasted with that of controls given no seizure prophylaxis. 5. The drug had no apparent effect on depth or duration of coma, but neither was the incidence of seizures reduced. 6. A controlled trial of prophylactic PB in young children with cerebral malaria is needed, but a larger dose than 10 mg kg-1 should be studied.