‘Prominent among the obligations of those who provide postgraduate education’, wrote Aubrey Lewis in 1961, ‘is the duty to recognize and further whatever talents for research their students possess’. He was largely responsible for introducing a system whereby those who wished to train at the Maudsley were expected to carry a modest research project as part of their DPM examination, although he was aware that ‘it is profitless, and can be unkind, to encourage competent clincians to attempt scientific investigations beyond their powers’. In the 14 years since his retirement, research by trainees has gone into a decline, so that even at the Maudsley only a minority are actively engaged in research. Crammer (1979) writes ‘People thought (research) would help them to a consultant post. Now they know it is not necessary. Training has become more formalized, with day-release courses and rotation schemes …’ Creed and Murray (1981) have made a detailed study of the views of Maudsley trainees about the way they were taught clinical skills; they have shown that between 1974 and 1977 there was a decline in the number of units which encouraged the trainee to undertake research, and that over the whole period only 27 per cent of the units gave trainees such encouragement. These investigators found that encouragement to do research was transmitted quite independently of other clinical teaching, so that it bore no relationship to variables such as the excellence of the academic instruction, feedback from the consultant, or workload.