A retrospective survey was conducted of 720 research protocols, approved by the Central Oxford Research Ethics Committee between 1984 and 1987, to determine the fate of research studies from inception. Forty-five per cent were clinical trials, 23% were observational studies and 32% were laboratory-based experimental studies. Further information was obtained on 487 studies, of which 287 (59%) had been completed, 100 (21%) had never started, 58 (12%) had been abandoned or were in abeyance and 42 (9%) were still ongoing, as of May 1990. Forty-three per cent of the original 487 studies were subsequently published or presented. The most frequent reason for not starting a study was failure to obtain funding (40%). The main reason for abandoning a study was difficulty in recruiting study participants (28%). Departure of one of the investigators from the institution and a variety of logistical problems were also common reasons for either not starting or abandoning a study. A thorough review of the pragmatic as well as the scientific aspects of a planned research project is important to minimize the initiation of studies that are unlikely to succeed.