A community survey of treatment regimens for acute diarrhoea in children was carried out in 10 villages in the Ona Ara Local Government Area of Oyo State, Nigeria, using a combination of qualitative (focus-group discussions) and quantitative (weekly surveillance of diarrhoea) methods. Focus-group discussions were conducted with parents of children aged less than 5 years, while a surveillance of diarrhoea among 550 children of same age was carried out during a 6-month period. The findings of the study showed that not all types of diarrhoea were recognized as illnesses, and only those considered to be illnesses were treated. Treatment often involved an adhoc group which comprised adults who were present at the time the illness occurred (including parents, neighbours, relatives, and elders). Certain beliefs and practices, such as associating types of diarrhoea with occupation or ethnic groups, categorizing the severity on perceived causes, and withholding certain foods during episodes of diarrhoea, were common factors in decision-making for seeking treatment. Antimicrobial agents were used in the case of 46.8% of 205 diarrhoeal episodes, and 28.5% were not at all treated. The usual practice of focusing on a target group, such as mothers, during educational interventions may need to be modified in communities where nearly every adult has a role in decision-making in relation to health. The need to adapt health policy and programmes to cultural norms should be addressed to improve the impact of programmes.