This paper provides a large-scale, rigorous evaluation of traffic calming projects in one U.S. city. The study area is New York City, which treated 391 streets with speed tables between 1996 and 2003. On the basis of crash frequencies for 5 years before treatment and 5 years after for treated streets and well-matched comparison streets, no evidence emerged that New York City's ambitious traffic calming program has led to a reduction in total crashes, pedestrian crashes, or injury crashes. This is in contrast to earlier, less carefully controlled evaluations that have reported significant reductions in crashes with traffic calming.