Clinical research on geriatric populations requires adaptation of traditional methodologies, modification of expectations, and the development of new procedures. Some aspects of research methodology are unique to studies of geriatric populations. Experience in this relatively new area of geriatric research indicates that new partnerships between researchers and long-term care providers are needed. These new relationships require an understanding of the needs of geriatric populations and of the differences between providers of long-term care and of acute-care. Researchers must consider heterogeneity of the population, the probability of multiple diagnoses and treatments, subject attrition, and the possibility of invalid data. Such considerations require extra staff, more time, and increased funding as well as new thinking about study design and protocol implementation.