This paper studies whether NAFTA contributed to the productivity convergence between Mexico and the US. Using data from the manufacturing sector for 1986 to 2000 and introducing a number of refinements in the computation of the total factor productivity and estimation methods, it shows that the productivity gap increased and converged to a higher gap level. Moreover, it finds no clear evidence for NAFTA's contribution to the TFP convergence. These findings are in a sharp contrast with that of the previous literature. The paper also finds that under NAFTA the industries which have smaller initial productivity gap experienced less increasing productivity gap than the industries which have larger initial productivity gap.