One of the difficulties in developing knowledge or evaluating actions in health and human services programs and practice is that the dominant research paradigm is reductionist and attempts to find cause--effect relations between discrete, quantifiable variables. While this paradigm makes important contributions to knowledge and action, there is growing argument in the practice and research communities that it is insufficient to make sense of what much of the ‘social work’ of health and human services is, and how its effects should be evaluated. At the same time, there are important counter-challenges that such practice is more ideological than theoretical, often little more than a series of normative claims. The "story/dialogue method" was developed to bridge this chasm. It was developed in a partnership between practitioners and researchers who were frustrated equally with researchers whose more conventional methods and assumptions often did not fit the "reality" of practice, and with practitioners who risked losing resources for their work by failing to articulate better practice-based theory.