Improving the accuracy of estimates of forest carbon exchange is a central priority for understanding ecosystem response to increased atmospheric CO2 levels and improving carbon cycle modelling. However, the spatially continuous parameterization of photosynthetic capacity (Vcmax) at global scales and appropriate temporal intervals within terrestrial biosphere models (TBMs) remains unresolved. This research investigates the use of biochemical parameters for modelling leaf photosynthetic capacity within a deciduous forest. Particular attention is given to the impacts of seasonality on both leaf biophysical variables and physiological processes, and their interdependent relationships. Four deciduous tree species were sampled across three growing seasons (2013–2015), approximately every 10 days for leaf chlorophyll content (ChlLeaf) and canopy structure. Leaf nitrogen (NArea) was also measured during 2014. Leaf photosynthesis was measured during 2014–2015 using a Li-6400 gas-exchange system, with A-Ci curves to model Vcmax. Results showed that seasonality and variations between species resulted in weak relationships between Vcmax normalized to 25°C (Vcmax25) and NArea (R2 = 0.62, P < 0.001), whereas ChlLeaf demonstrated a much stronger correlation with Vcmax25 (R2 = 0.78, P < 0.001). The relationship between ChlLeaf and NArea was also weak (R2 = 0.47, P < 0.001), possibly due to the dynamic partitioning of nitrogen, between and within photosynthetic and nonphotosynthetic fractions. The spatial and temporal variability of Vcmax25 was mapped using Landsat TM/ETM satellite data across the forest site, using physical models to derive ChlLeaf. TBMs largely treat photosynthetic parameters as either fixed constants or varying according to leaf nitrogen content. This research challenges assumptions that simple NArea–Vcmax25 relationships can reliably be used to constrain photosynthetic capacity in TBMs, even within the same plant functional type. It is suggested that ChlLeaf provides a more accurate, direct proxy for Vcmax25 and is also more easily retrievable from satellite data. These results have important implications for carbon modelling within deciduous ecosystems.