Test beds have become an integral part of the weather enterprise, bridging research and forecast services by transitioning innovative tools and tested methods that impact forecasts and forecast users. O ver roughly the last decade, a variety of "test beds" have come into existence focused on high-impact weather and the core tools of meteorology-observations, models, and fundamental understanding of the underlying physical processes. They have entered the proverbial "valley of death" between research and forecast operations (NAS 2000), Develop and introduce new ideas, data, etc. Input Revise and iterate Experiment and demonstrate End testing Output Test and refine loop V Assess impacts and evaluate and have survived. This paper provides a brief background on how this happened; summarizes test bed origins, methods, and selected accomplishments; and provides a perspective on the future of test beds in our field. Dabbert et al. (2005) provides a useful description of test beds from early in their development and Fig.