In developing countries, rural water supply systems, generally planned for a single purpose, are used for domestic and productive activities. Research has found linkages between access to water and livelihoods, thus, promoting the implementation of Multiple Uses of Water Systems (MUS). However, there is concern on how users with more productive assets may exploit the highest levels of service promoted by MUS affecting users with fewer chances to use water productively. This article analyses through a case study and using multivariate techniques (Principal Component Analysis and Cluster Analysis): key variables to characterize the users of a de facto MUS; identifies groups of users according to those variables, and discusses implications of the findings in the design and management of systems. Results show correlations for the variables Homestead Area, Storage Capacity and Number of Cows, with Household Water Consumption. Two groups of customers with statistically significant differences for key variables are identified. The results suggest that rural water systems should not be designed with household size as the only criterion. It also highlights the need of managerial rules for equitable access to wáter in systems, recognizing diversity of users, livelihoods and thus, water demands.