The study of minority ethnic older people in Britain is hampered by the lack of a reliable sampling frame. There are few national studies and most research projects are small and involve localised samples. The infrastructure for doing research with such population groups is poorly set up. This article attempts to address the needs of practitioners by reviewing the use of common sampling frames and approaches, pointing out caveats, and highlighting potential. More recent and innovative sampling approaches are also discussed. The article deals with issues of bias, reliability, perspective and geography. Recommendations for improved practice include being more transparent in reporting the derivation of samples, making more efficient use of existing data sets, better ways of working with community organisations as well as more effective collaborative research. It is also argued that various authorities and funding bodies should increase investment in this area, especially given the projected absolute and proportionate increases in minority ethnic older people in Britain.