There is significant documentation in the literature of barriers that may prevent research results from being utilized by agencies to inform and impact practice and policy. Such barriers pertain to several factors as follows: (a) those related to the nature of the research enterprise itself, (b) those related to differences between the producers and consumers of research, and (c) barriers arising from the differences in organizational contexts of researchers and case management and supported employment agency staff. This article discusses a collaborative relationship between university researchers and agency practitioners in the context of a research project studying the implementation of supported employment, an evidence-based practice. As a case example, it provides an exemplar of the problems and issues of conducting mental health research with community-based agencies and offers strategies and case examples that address these issues. Keywords: case management; collaboration; university-agency research; mental health; supported employment This article discusses the collaborative relationship between university researchers and agency practitioners in the conduct of a research project studying the implementation of supported employment (SE), an evidence-based practice. As a case example, it provides an exemplar of the problems and issues of conducting mental health research with community-based agencies and offers strategies and case examples that address these issues. Human service agencies in general and mental health agencies in particular are under increasing pressure from policymakers and funders to demonstrate the effectiveness of their services. Over time, we have seen a shift from a focus on inputs, such as services provided by mental health case managers, to examinations of the effects of those services on mental health consumers and their families. Thus, research studies that gather empirical data to document the effectiveness of services can be expected to play an expanding role in the mental health and other human service sectors in the future. However, there is significant documentation in the literature of barriers that may prevent research results from being utilized by agencies to inform and impact practice and policy. Such barriers pertain to several factors: (a) those related to the nature of the research enterprise itself, (b) those related to differences between the producers and consumers of research, and (c) barriers arising from the differences in organizational contexts of researchers and agency staff. First, several characteristics of the research enterprise are potential barriers to the application of findings to practice and policy (Roth & Fonagy, 1996). Limitations imposed by research designs often restrict the usefulness of the results. For instance, methods employed in controlled studies may not generalize to "real-world" settings, thus creating a lack of fit between the research project and agency (Fuller, Caldwell, & Allen, 1992; Hellervik, Hasucha, & Schneider, 1992; Subramanian, Siegel, & Garcia, 1994). Furthermore, research findings are typically expressed in terms of statistical significance. Whether this translates into clinical significance is seldomly addressed by the researcher and often is not self-evident to practitioners, such as mental health case managers. Thus, agency staff may not feel that research findings are relevant to the real world of their agencies' work (Blum, Biegel, Tracy, & Cole, 1995; Garland, Plemmons, & Koontz, 2006). Conducting research is a time-consuming, labor-intensive process. Only recently has there been increased attention to the need for the dissemination of the study findings as an expected component of applied research studies and for the need to budget for costs of such dissemination in terms of either dollars or staff time. Thus, agency practitioners have reported that they do not have access to research data and are unaware of research findings, that the data may have been presented in a way that was not useful to them, or that research findings were not available at a time when agencies needed to make practice or service delivery decisions (Reback, Cohen, Freese, & Shoptaw, 2002; Welsh & Zajac, 2004). …