As part of a longitudinal, community-based study of diarrhoeal morbidity in a peri-urban community in Lima, Peru, a household survey was administered to ascertain possible risk factors, based on transmission routes, for diarrhoeal incidence. Socioeconomic information was also obtained in the survey and a composite socioeconomic status (SES) indicator was created based on four variables: income (wealth), ownership of 4 functioning electrical household appliances, community participation, and house construction. Both transmission factors and the SES indicator were analysed for their effects on diarrhoeal incidence using both bivariate and multivariate methods. The SES indicator, method of water storage, if the child was seen eating faeces or soil were all significantly associated with diarrhoeal incidence. In a final logistic model, water storage, location of defecation for children, child eating soil or faeces, and age, demonstrated significant results. Children in households with water stored in containers without a faucet were twice as likely to have a high incidence of diarrhoea (greater than 7 episodes/child/year). The SES indicator was not significant in the logistic model, but high SES was associated with whether or not the child was reported as having been seen eating faeces or soil and with non-use of latrines by adults. Also low SES households were more likely to have better water storage methods. Therefore, it would seem that (SES) does not independently determine diarrhoeal incidence, but rather may be functioning through these transmission factors to affect diarrhoeal incidence.