The increase in drop-out rates, especially among older people, in longitudinal studies is a matter for concern if the results are to be valid. The research reported here contains a number of pieces of evidence that might help address the problem. These include a literature review, a survey of some longitudinal studies, secondary analysis of the data from a longitudinal study of civil servants, and new data from focus groups and telephone interviews with participants from this current study. The findings (...) relate to reasons why older people continue to take part in such studies, why they drop out and what might encourage them to continue participating. The main reasons for drop out relate to being older, having cognitive impairment and poor functioning, living alone, being a woman and being less educated. There were also a range of practical reasons including the study being too time-consuming and problems with transport. (shrink)