Meaningful and clinically relevant evaluation of psychoanalytical psychotherapy requires new models for research. Such evaluation needs to take account of the complexities of personality and inner-world change as well as the more simply assessed changes in external behaviour.

A pilot study to develop a suitable methodology is described. These preliminary results are on a consecutive series of twenty in care and adopted children entering psychotherapy at the Tavistock Clinic. The project evolved from a previous study of psychotherapy with severely deprived children (Boston & Szur 1983), but the method could be used with any group of children in psychoanalytic psychotherapy, and possibly also, in an adapted form, with adults.

Therapists' initial aims and predictions are compared with later assessments of both external, and a wide range of internal, changes. Methods of validation are discussed and case-illustrations given.