Sir: The ability to elicit a history and to examine the mental state are essential clinical skills, but ones that are difficult to test. Direct observation would appear to be the most appropriate way of assessing such skills. However, Dale (Psychiatric Bulletin, October 2000, 24, 395) is concerned about the validity of the objective structured clinical examination (OSCE) in the revised MRCPsych examination (Katona et al, Psychiatric Bulletin, July 2000, 24, 276-278).

Currently both parts of the examination include the assessment of a single unobserved long case. This has doubtful reliability and provides an insufficient variety of clinical situations. In contrast, the OSCE is designed to test the candidates' competence in a range of clinical and practical skills.

At St George's Hospital the incorporation of an OSCE into undergraduate assessment has been a valuable learning experience for the examiners. It has indicated deficits in clinical teaching. In particular, we have reflected on how little observation of assessment by the students takes place during their psychiatry attachments.

It is pessimistic to suggest that the OSCE will delay the development of postgraduate clinical skills. Instead, the exam will emphasise their importance. Candidates preparing for the exam would be well advised to request that their education supervisors observe them carrying out patient assessments.

The OSCE will improve assessment in the MRCPsych examination and may provide the impetus to improve clinical teaching in the ‘real world’.