Many cities throughout the United States require developers of new residential construction to provide a minimum number of accompanying off-street parking spaces. However, critics argue that these requirements increase housing costs by bundling an oversupply of parking with new housing and by reducing the number of units developers could otherwise fit on a given lot. Furthermore, the requirements reduce the subsequent direct costs of car ownership by forcing up-front, or subsidizing, consumption of parking spaces, which leads to increases in auto-use and its related externalities. Such critics advocate eliminating or at least reducing the requirements to be more responsive to locational context, particularly proximity to transit. In this article, we explore the theoretical objections to minimum parking requirements and the limited empirical literature. We then use lot-level data and GIS to analyze parking requirements in New York City to determine to what extent they are already effectively sensitive to transit proximity. Finally, we examine developer response to parking requirements by comparing the number of spaces that are actually built to the number required by applicable zoning law. Our results indicate that the per-unit parking requirement in New York is, on average, lower in areas near rail transit stations, but the required number of spaces per square foot of lot area is higher, on average, in transit accessible areas. We also find that by and large, developers tend to build only the bare minimum of parking required by zoning, suggesting that the minimum parking requirements are binding for developers, as argued by critics, and that developers do not simply build parking out of perceived marked need. Our results raise the possibility that even in cities with complex and tailored parking requirements, there is room to tie the requirements more closely to contextual factors. Further, such changes are likely to result in fewer parking spaces from residential developers.