This paper offers a critique of existing methods of research on psychoanalytic treatment and proposes audio-recording in order (i) to improve the data record, and (ii) to facilitate the separation of therapeutic from true research activities. Effects of research upon patient and analyst are considered and the objections to recording re-evaluated. We conclude that recorded research analysis needs to be done, that the crucial problems may lie in the therapist as much as, if not more than, in the patient, and we

believe that such an analysis can be successfully performed. If not, that demonstration in itself will constitute a valuable addition to our fund of information about psychoanalytic treatment and provide other leads for future psychoanalytic research.