Community-based participatory research strategies are used at Montana State University (MSU) to co-create research data with Native American and rural communities in the American West. To increase benefit and minimize harm to participating communities, the MSU Library has partnered with the NIH-funded Center for American Indian and Rural Health Equity (CAIRHE) to create metadata records for restricted data, thus making the data available to community members while maintaining the confidentiality and sovereignty of the data. This project can serve as a model for other libraries wishing to build similar partnerships. The Library partnership with CAIRHE addresses four challenges: 1. Sensitive data. The data relate to sensitive topics in community health and they are co-created with small communities, which makes deidentification difficult. 2. Data sovereignty. In accordance with Native American data sovereignty principles, data belong to communities where research is conducted. This means that publishing data in a repository with restricted access may not an appropriate solution—doing so may put the data into the hands of institutions with whom the communities do not have relationships. 3. Trusted data storage. It is common that small, rural communities in [our state] don’t have the resources to store and preserve data locally. CAIRHE is trusted by its partner communities, so storing data using MSU infrastructure is the best solution in this particular case. However, appropriate members of communities must be able to access the data. 4. In-progress data. In-progress data need to be available to communities to support time-sensitive community efforts such as grant proposals and public health initiatives. This lightning talk provides an overview of the process of collaborating with CAIRHE and customizing MSU’s data catalog, including outreach, communication, curation strategies, and lessons learned. Collaborations like the one at MSU can help members of communities who co-create research to find, request, and access appropriate data.