Despite a debate lasting some decades, there has been little or no clear resolution about the way knowledge may be related to practice. To a considerable degree this is because of an assumption that knowledge is ‘knowledge as product’: given knowledge, researched and applied to practice. In recent years, the concept of ‘knowledge as process’ has emerged, focusing on knowledge of rigorous ways to think about practice situations. This has generated a focus on hypothesis generation, testing and falsification, as well as lessons to be learned from rigorous methods of qualitative analysis in social research. Rigorous social workers, it has been suggested, would behave like ‘practical qualitative researchers of the individual instant’. However, there is little empirical research on the matter. This paper evaluates process knowledge forms used by social workers. It presents findings that social workers vary considerably in the extent to which they develop hypotheses about the whole case, or aspects of it, in the depth with which they analyse cases, and the systematic way they pursue an information gathering strategy. The paper concludes that there is clear evidence of social workers behaving as practical qualitative researchers, but the rigour with which they do so varies, with fundamental implications for the training process, and the extent to which rigorous strategies, along the lines of research methodology, should be taught.