In many developing countries, lack of access to water and its unsustainable use and adverse health impacts are important policy issues. The challenge of improving water services in developing countries is not only limited to the provision of new connections for non-piped-source-dependent households but also to improving the service level for the already connected households to meet peoples' needs. We conducted an empirical study in Sri Lanka on 307 households with piped water supply. A choice experiment was used to assess the potential welfare gains from alternative water supply schemes in terms of water quality, reliability, and pressure, and how the schemes are managed. Currently, over 20 different water supply systems managed by either governmental organizations or community-based organizations (CBOs) provide piped water supply. We found that households are willing to pay a substantial amount for improved water service, particularly for water quality improvements. Using the latent class and random parameter logit models, we found evidence of heterogeneity in the preferences for water service improvements. We suggest that the heterogeneous preferences of the household groups should be considered in the cost–benefit analysis to pave a way for rational policy-making on water service improvements.