The San Francisco Bay Area in California, USA is a highly heterogeneous region in climate, topography, and habitats, as well as in its political and economic interests. Successful conservation strategies must consider various current and future competing demands for the land, and should pay special attention to livestock grazing, the dominant non-urban land-use. The main objective of this study was to predict changes in rangeland forage production in response to changes in temperature and precipitation projected by downscaled output from global climate models. Daily temperature and precipitation data generated by four climate models were used as input variables for an existing rangeland forage production model (linear regression) for California's annual rangelands and projected on 244 12 km x 12 km grid cells for eight Bay Area counties. Climate model projections suggest that forage production in Bay Area rangelands may be enhanced by future conditions in most years, at least in terms of peak standing crop. However, the timing of production is as important as its peak, and altered precipitation patterns could mean delayed germination, resulting in shorter growing seasons and longer periods of inadequate forage quality. An increase in the frequency of extremely dry years also increases the uncertainty of forage availability. These shifts in forage production will affect the economic viability and conservation strategies for rangelands in the San Francisco Bay Area.