This article reviews the empirical data obtained in existing studies on the multimodal treatment of chronic pain. The majority of these 'treatment packages' are based on a cognitive/behavioural perspective. The articles for review were selected from the scientific literature on this subject which has appeared since the first publication of Fordyce in 1973. The following aspects have been analysed: the goal and structure of the treatment programmes; the method of evaluating treatment results; and the indications for treatment. Programmes for in-patients and out-patients have been compared because it is very likely that there are differences between the treatment methods and study populations. The interval validity and clinical relevance of the available research are discussed in the conclusion.