Modeling stomatal behavior is critical in research on land–atmosphere interactions and climate change. The most common model uses an existing relationship between photosynthesis and stomatal conductance. However, its parameters have been determined using infrequent and leaf-scale gas-exchange measurements and may not be representative of the whole canopy in time and space. In this study, we used a top-down approach based on a double-source canopy model and eddy flux measurements throughout the growing season. Using this approach, we quantified the canopy-scale relationship between gross photosynthesis and stomatal conductance for 3 years and their relationships with leaf nitrogen content throughout each growing season above a paddy rice canopy in Japan. The canopy-averaged stomatal conductance (gsc) increased with increasing gross photosynthesis per unit green leaf area (Ag), as was the case with leaf-scale measurements, and 41–90% of its variation was explained by variations in Ag adjusted to account for the leaf-to-air vapor-pressure deficit and CO2 concentration using the Leuning model. The slope (m) in this model (gsc versus the adjusted Ag) was almost constant within a 15-day period, but changed seasonally. The m values determined using an ensemble dataset for two mid-growing-season 15-day periods were 30.8 (SE = 0.5), 29.9 (SE = 0.7), and 29.9 (SE = 0.6) in 2004, 2005, and 2006, respectively; the overall mid-season value was 30.3 and did not greatly differ among the 3 years. However, m appeared to be higher during the early and late growing seasons. The ontogenic changes in leaf nitrogen content strongly affected Ag and thus gsc. In addition, we have discussed the agronomic impacts of the interactions between leaf nitrogen content and gsc. Despite limitations in the observations and modeling, our canopy-scale results emphasize the importance of continuous, season-long estimates of stomatal model parameters for crops using top-down approaches.