This paper uses a structural vector autoregression (SVAR) to model the US natural gas market. Domestic natural gas production has dramatically increased in the late 2000’s as producers implemented improved horizontal drilling and hydraulic fracturing techniques to extract natural gas from shale deposits. This paper applies techniques developed in the crude-oil literature that incorporate time-varying parameters (TVP), which more realistically model a changing market. Results suggest the supply curve associated with aggregate demand shocks, residential demand shocks, and other natural gas market demand shocks has become more elastic during the latter part of the sample. This latter period corresponds to the recent rapid expansion in shale gas production. By contrast, the curvature of the supply curve associated with precautionary inventory demand shocks has remained at a stable level for most of the sample since 2000, and has not changed significantly even with the rapid growth of shale gas production. Furthermore, it appears that in the new era of ample natural gas supply, the demand curve has also become more elastic, partly reflecting greater flexibility in fuel use substitution.