Most empirical studies for the post Bretton‐Woods period fail to find evidence of a long‐run Purchasing Power Parity (PPP) relationship. An investigation into the failure of PPP is made in this study by using disaggregated price data. This disaggregation is on two levels: location (prices from US and Canadian cities rather than national aggregates) and type of goods (e.g., fuel oil, a tradable commodity and local public transportation, a non‐tradable). This disaggregation allows for the testing of the importance of borders (implying an exchange rate), distances, and types of goods in the failure of PPP. The analysis conducted suggests that both country borders and distances play a significant role. However, there is mixed evidence concerning type of goods as an important determinant of the failure of PPP.