Partial and total measures of factor productivity are presented for the pulp and paper industries of the United States and Canada, from 1959 to 1987. Total factor productivity was measured with (1) a Tornqvist–Theil index, (2) a nonparametric index with translating hypothesis, and (3) a nonparametric index with distance functions. Method 1 implied a constant return to scale translog production function. Methods 2 and 3 removed any assumption on the functional form of the production function. Furthermore, method 3 allowed for fully disaggregated outputs. Methods 1 and 3 gave similar results within countries: an increase in total factor productivity of 0.7% per year in the United States and of 0.5% per year in Canada. Method 2 gave rates of growth of total factor productivity that were twice as high, but unreliable because of the assumptions of the method. From 1961 to 1984, when comparable data are available, methods 1 and 3 gave growth rates of total factor productivity that were significantly higher, statistically, in the United States than in Canada. Nevertheless, the differences seem to be too small to be of economic significance.