Cost-effectiveness of anti-malaria activities in Sri Lanka was examined from a societal point of view using Matale district as a case study. Intermediate level cost-effectiveness indicators were measured for curative care and preventive care separately. In measuring cost-effectiveness indicators, community acceptance of and compliance with control measures were incorporated into the analysis particularly for preventive measures with a view of focusing on community effectiveness of control measures and strategies. Based on community perceptions, the household level illness rate of malaria (HLIR) was used to measure effectiveness of the combinations of preventive and curative measures; it was an attempt to widen the scope of the cost-effectiveness analysis in its application to ongoing programmes in the absence of a control area or randomized control trials. The main preventive measure of residual spraying was found to be a waste of resources because there was low acceptance and no evidence that those accepting the spraying subsequently had a lower prevalence rate. Knowledge, attitudes, practices and socio-economic factors underlying community behaviour were analyzed particularly to identify policy options to increase the efficiency of the control programme.S elf preventive measures were common among all community groups. The community demonstrated a high sensitivity towards illness and considerable use of the private sector for curative care. However the public sector tended to be the flnal resort for severe illness, irrespective of socio-economic background. Supply deflciencies were a prime cause for the tendency towards the private sector. The cost implications and possible outputs of two alternative policy options were examined in detail. Incorporation of community behaviour into the cost-effectiveness analysis to assess community effectiveness should urge economists and epidemiologists to work together, particularly in assessing the validity of community perceptions of illness prior to undertaking an economic evaluation.