The aim of this retrospective study was to determine the outcome of orthodontic treatment carried out on patients by postgraduate students at the Katholieke Universiteit Leuven, Belgium. The treatment outcome of 292 'final examination' patients and of 287 'control' patients was compared by means of the Peer Assessment Rating (PAR) index. The sample consisted of dental casts representing a wide range of malocclusions at the start of treatment and post-treatment. All patients received non-surgical treatment between 1987 and 1996 by one of 18 different postgraduate students. The data were analysed with a variant of the analysis of covariance. A significantly higher (P < 0.001) treatment standard was found for final examination patients compared to the control, indicated by the mean percentage PAR score reduction of 79.1% and 70.7%, respectively. When the results are expressed in terms of treatment outcome, 44.5% of the examination patients and 44.0% of the controls were allocated to the 'Greatly improved' group, while 3.1% of the patients examined and 7.3% of the control patients were classified as 'Worse or no different'.