A spatial Durbin model (SDM) is developed to estimate the open space proximity values in Fresno-Clovis metropolitan area in Central California circa 2005 to 2012, a time period covering a full market cycle with distinctive boom, bust, and recovery market phases. By controlling for distressed market conditions (vacant, foreclosure, and short-sale transactions) and spillover pricing effects, we find significant but heterogeneous values for proximity to five types of open space amenities. Furthermore, proximity values for parks and golf courses are found to drop by more than 20 % in the bust period relative to the boom period. Results show that failing to capture distressed market conditions and market phases, when the housing market is out of equilibrium, leads to biased results. In addition, our findings suggest that ignoring spatial dependence and spillover pricing effects presents a potential bias in estimating open space amenity values. The SDM, which considers spatial dependence and spillovers at a finer scale in the neighborhood, outperforms region fixed effects models when spatial spillovers are significant.