The standard measure of productivity growth is the solow residual. Its evaluation requires data on factor input shares or prices. Since these prices are presumed to match factor productivities, the standard procedure amounts to accepting at face value what is supposed to be measured. In this paper we estimate total factor productivity growth without recourse to data on factor input prices. Factor productivities are defined as lagrange multipliers to the program that maximizes the level of domestic final demand. The consequent measure of total factor productivity is shown to encompass not only the solow residual, but also the efficiency change of frontier analysis and the hitherto slippery terms-of-trade effect. Using input-output tables from 1962 to 1991 we show that the source of canadian productivity growth has shifted from technical change to terms-of-trade effects.