In‐service training within social service agencies is recognized as a key means through which staff are provided with the necessary knowledge and skills to improve overall agency performance and achieve the objectives of social policy. Furthermore, reports of training expenditure within social services departments in the UK suggest increasing amounts of monies are invested in such activity in order to meet the changing demands placed on social care and greater expectations for higher standards in service delivery. Yet to what extent is the faith placed in much of this training by social service agencies actually justified? In short, how much do we know about the actual impact of in‐service training within social service organizations in terms of it achieving the aims set by policy makers? This article attempts to answer this question in order to assist in the development of a far more empirically based framework for understanding training processes within social service agencies. A review of the literature for studies published between 1974 and 1997 detailing an evaluation of in‐service training programmes within social service agencies identified only 20 such studies. Problems associated with both the evaluative criteria utilized and the research methodologies employed in these studies meant that in many instances conclusions regarding the actual impact of training could only be tentatively judged. As a result, an analysis of these studies found that although training may have an impact on trainees in terms of satisfaction or knowledge gain, results regarding impact on behaviour are far more inconclusive. In addition it is not at all certain that such training will necessarily result in changes in performance back in the workplace. The findings from the review suggest far more research is required of sufficient rigour to underpin our knowledge in this important area.