Healthcare professionals have become increasingly concerned with evaluating the impact of their interventions. Consumerism, quality initiatives and financial constraints have contributed in large measure to this concern. Consequently, the focus of many studies which involve the introduction of new and untried techniques is often on demonstrating their relative usefulness. This study sought to evaluate the usefulness and acceptability of computer assisted learning (CAL) for use in the education of renal patients on continuous ambulatory peritoneal dialysis. However, several problems were encountered which limited or impaired the evaluation process. We describe these problems and use them to illustrate the difficulties inherent in evaluative research.