An intervention study was developed from risk-factors associated with faeco-oral transmission, based on the levels of contamination in stored water and fingertip-rinses from households in rural north-east Thailand. This was designed to improve: (a) handwashing, particularly before cooking/eating and after defecation: (b) washing dishes immediately after use. Verbal messages were administered to two intervention groups, one also received a plastic container with a tap to assist these activities. Indicators of compliance were the direct observation of soaking dishes and the presence of faecal streptococci from fingertip-rinses; the main outcome indicator was Escherichia coli contamination of stored water. The intervention group receiving the container was significantly better than the control for indicators of compliance (P less than 0.001 and P less than 0.01) and its stored water was significantly less contaminated (P less than 0.001). There was no significant improvement to the other intervention group, although some features of the intervention had clearly been made available to the control group. Humidity was significantly correlated with fingertip contamination (r = 0.2; P less than 0.001) and with the peak of reported diarrhoea around the beginning of the rainy season.