While the literature is clear that areas endowed with scenic beauty, lakes, forests, and wildlife among other natural and built amenities are experiencing robust economic growth, the theoretical foundations for this work are weak. In this study, we employ a Bayesian Modeling Average (BMA) approach to address the problem of model specification. Using data for US counties, we estimate a neoclassical growth model while looking at growth over the 1990s. We look at three separate ways of measuring amenities and recreational opportunities that build on those amenities. While the results suggest that higher amenity areas experience faster growth, still some level of value-added development may be required to realize that growth.