The paper builds a unique industry‐level panel data set to estimate the border effects associated with US–Canada trade for each year from 1992 to 2005. We first establish the theoretical foundation of our empirical model as a multisector version of Anderson and van Wincoop. Estimates from data aggregated at the province/state level yield border effects that increase slightly in the early 1990s, then decline after the implementation of the North American Free Trade Agreement (NAFTA), but then increase significantly after 2001. Results based on three‐digit NAICS level data reveal higher border effects in the early 1990s and substantial heterogeneity across industries. The results are robust to a variety of specifications and models, and they suggest that the security measures adopted in the aftermath of the tragic events of 11 September 2001 had considerable adverse effects on US–Canada trade.