Trade agreements do not necessitate business cycle comovement. Focusing on NAFTA, we investigate whether business cycles in Canada, Mexico, and the US have become more synchronous after the landmark trade agreement came into effect in 1994. To this end, using the newly-developed Hamilton filter, we decompose the real GDPs of the three countries to derive their business cycle components; then, we conduct time-difference analyses, which illuminate correlations at different time intervals, to study business cycle synchronization. We find that business cycles in Mexico and the US have become positively correlated after NAFTA—they were weakly and negatively correlated during the pre-NAFTA period. Contrastingly, correlations amongst the US and Canadian business cycles have weakened during the post-NAFTA period; nevertheless, these two countries' business cycles continue to be tightly and positively correlated. The oft-used Hodrick-Prescott filter is utilized to confirm the robustness of the results—the two filters lead to similar conclusions.