A bottom-up and a top-down model were used to estimate the effect of temperature on monthly gross primary productivity (GPP) of sugar maple (Acer saccharum Marsh.). The bottom-up model computed canopy photosynthesis at an hourly time step from detailed physiological sub-models of leaf photosynthesis and stomatal conductance. Leaf mass per area was used as a covariable to integrate photosynthesis through the canopy. The top-down model used a radiation-use efficiency coefficient to relate canopy gross photosynthesis to absorbed photosynthetically active radiation at a monthly time step. The parameters of the top-down model were estimated from simulations with the bottom-up model. Forty single-year simulations were made using records of daily maximum and minimum temperatures from weather stations selected within the natural range of sugar maple in the province of Québec, Canada. Leaf area index was randomly varied between 4 and 10. Within a broad range of values, temperature had a minor effect on predicted monthly canopy-level GPP and its contribution to explaining the variability of GPP was low, both through its direct effect on photosynthetic processes (1.1%), and indirectly through the effect of relative humidity on stomatal conductance (4.0%). This result was unchanged when key parameters relating photosynthesis to temperature and stomatal conductance to atmospheric humidity were changed in the bottom-up model. An increase in time step from hourly to monthly resulted in a downward shift in the optimum temperature range for photosynthesis, from 30 degrees C for a leaf at saturating irradiance to 22 degrees C for the canopy at a monthly time scale.