Hourly plot-level transpiration measurements were carried out in a stand of sugar maple (Acer saccharum Marsh.) near Quebec Coty, Canada, durng one summer using radial flow meters installed in selected trees. The measurements were used to validate transpiration estimates obtained from a multi-layer gas exchange model that included the modeling of the within-canopy radiation and the thermal balance of the leaves. The comparison between modeled and measured hourly transpiration showed no bias in the predictious, and an r2 value of 0.80. Because of the coupling between transpiration and photosynthesis built in the model, these results suggest that modeled estimates of photosynthesis should also be well-related to actual rates at the stand-level. The transpiration data were also used to adjust an empirical transpiration model based on the Penman-Montheith equation in which the canopy conductance term was replaced by a function of global radiation, vapor pressure deficit and soil water depletion. This empirically-fitted model captured 85% of the variation observed in the data, including the effect of soil water depletion during a late-season drought. The soil water modifier included within this equation was compared with other soil water modifier functions obtained from the literature. The comparison highlights two difficulties in the derivation of soil water modifier applicable outside experimental areas. The first is in the determination of rooting depth so taht mass balances of soil water content can be carried out. The second is in the determination of soil physical properties so that absolute values of moisture contents can be translated int o relatives values of water availability.