While infrastructure conditions constitute ‘primary routes’, contamination of water within households and other behavioural determinants are considered as ‘secondary routes’. However, recontaminated water has been considered not to constitute a serious risk though it occurs commonly in poorer societies. A study was conducted in Delhi where individual risk factors were located within a larger socio-economic, political and administrative framework, as they were often independent variables. This component of the larger study hypothesised that behavioural factors at individual household levels lose significance as major determinants of diarrhoeal diseases once they are analysed in a holistic epidemiology frame. Determinants at the household level were explored through a dataset based on a primary survey of 300 households in three slum clusters. Amongst households storing municipal water (proven to be safe at source), adhering to the best storage practices did not translate into lower incidence rates as compared to those with relatively unsafe practices. The explanation lay in factors which were external to the home and beyond the control of the affected household. Thus, household level behavioural factors such as storage practises should not be analysed in isolation as determinants of diarrhoeal illness particularly when pitted against stronger neighbourhood and external determinants.