A survey was done in June 1983 in Saradidi, Kenya, one year after the inception of a community-based malaria control programme to determine if people were obtaining malaria treatment from volunteer village health helpers (VHH's) chosen by the community. Ten of 36 villages were randomly chosen. From these ten villages, 100 households were randomly selected and 222 people ten years of age or more were interviewed: 113 (50.9%) had a history of malaria in the previous two weeks and 82 (72.6% of 113) had taken medicine for malaria in that period. Of these 82, 51.2% obtained drug from the VHH, 28% purchased it from a shop, 12.2% from a health facility, 4.9% from family members and 3.7% from a private practitioner or a shop outside Saradidi. Reasons given for not obtaining treatment from the VHH's among the 40 people who went elsewhere for treatment included: the VHH was not at home when needed (35%); the VHH had no drugs (22.5%); the patient was too sick for the VHH to treat (10%); had drugs already in the home (10%); ‘not registered’ with VHH (10%); VHH ‘no good’ (7.5%); and more ‘convenient’ to go elsewhere (5%). Similar results found on questioning the mother were obtained for 103 children under nine years old in these households; 67 (65.0%) children had a history of malaria in the previous two weeks and 59 (88.1%) of these 67 children had received antimalarial treatment. The VHH was the principal source of treatment (50.8% of 59), followed by health facilities (20.3%) and shops (18.6%). Again, the major reasons for not obtaining treatment from the VHH was that the VHH was not at home (23.3% of 30), the VHH had no drug (13.3%), or the patient was too sick (20%). It is clear that the community-based programme was the main source of antimalarial treatment in Saradidi. Drug consumption was high in both adults and children, perhaps because the survey was done during the season of peak malaria transmission. The main reasons for failure to obtain treatment from the VHH were logistical or organizational.