Obviously, evaluations of parenting programmes are important for improving clinical practice. Such evaluation studies can show us whether a parenting programme works and for whom and under which circumstances the programme has the best effects. An additional, often neglected, role of evaluation studies is to provide a test of the theory on which the programme is based. In the present paper, we discuss the ways in which evaluation studies can be used to test and refine theories pertaining to parenting behaviour and child development. Several examples of such studies are presented and both conceptual (specifying the theory-based expectations regarding working mechanisms) and methodological issues (quality criteria for real-life evaluation studies of parenting programmes) involved in evaluation research are addressed.