Researchers and policy-makers interested in the needs of the frail elderly in community and institutional settings have had limited success in forming policy based on empirical evidence. The reasons for this have included a variety of organizational and political considerations (e.g., lack of effective communication mechanisms), conceptual and methodological issues (e.g., problems with reliability and validity of existing data sources) and limitations in knowledge (e.g., lack of longitudinal data across health care sectors). The Minimum Data Set (MDS) series of instruments may prove useful in dealing with these difficulties. The MDS instruments have multiple uses for different audiences (e.g., care provision, funding and quality improvement). Extensive international testing has shown the MDS to be valid and reliable, and translations are available in 11 different languages. As implementation of the MDS begins in Canada, a number of new concerns will need to be addressed (e.g., confidentiality and access to data).