A dynamic model of leaf CO₂ assimilation was developed as an extension of the canonical steady-state model, by adding the effects of energy-dependent non-photochemical quenching (qE), chloroplast movement, photoinhibition, regulation of enzyme activity in the Calvin cycle, metabolite concentrations, and dynamic CO₂ diffusion. The model was calibrated and tested successfully using published measurements of gas exchange and chlorophyll fluorescence on Arabidopsis thaliana ecotype Col-0 and several photosynthetic mutants and transformants affecting the regulation of Rubisco activity (rca-2 and rwt43), non-photochemical quenching (npq4-1 and npq1-2), and sucrose synthesis (spsa1). The potential improvements on CO₂ assimilation under fluctuating irradiance that can be achieved by removing the kinetic limitations on the regulation of enzyme activities, electron transport, and stomatal conductance were calculated in silico for different scenarios. The model predicted that the rates of activation of enzymes in the Calvin cycle and stomatal opening were the most limiting (up to 17% improvement) and that effects varied with the frequency of fluctuations. On the other hand, relaxation of qE and chloroplast movement had a strong effect on average low-irradiance CO₂ assimilation (up to 10% improvement). Strong synergies among processes were found, such that removing all kinetic limitations simultaneously resulted in improvements of up to 32%.