Rubisco is the most abundant protein on Earth that serves as the primary engine of carbon assimilation. It is characterized by a slow rate and low specificity for CO2 leading to photorespiration. We analyze here the challenges of operation of this enzyme as the main carbon fixation engine. The high concentration of Rubisco exceeds that of its substrate CO2 by 2–3 orders of magnitude; however, the total pool of available carbon in chloroplast, i.e., mainly bicarbonate, is comparable to the concentration of Rubisco active sites. This makes the reactant stationary assumption (RSA), which is essential as a condition of satisfying the Michaelis–Menten (MM) kinetics, valid if we assume that the delivery of CO2 from this pool is not limiting. The RSA is supported by active carbonic anhydrases (CA) that quickly equilibrate bicarbonate and CO2 pools and supply CO2 to Rubisco. While the operation of stromal CA is independent of light reactions, the thylakoidal CA associated with PSII and pumping CO2 from the thylakoid lumen is coordinated with the rate of electron transport, water splitting and proton gradient across the thylakoid membrane. At high CO2 concentrations, CA becomes less efficient (the equilibrium becomes unfavorable), so a deviation from the MM kinetics is observed, consistent with Rubisco reaching its Vmax at approximately 50% lower level than expected from the classical MM curve. Previously, this deviation was controversially explained by the limitation of RuBP regeneration. At low ambient CO2 and correspondingly limited capacity of the bicarbonate pool, its depletion at Rubisco sites is relieved in that the enzyme utilizes O2 instead of CO2, i.e., by photorespiration. In this process, CO2 is supplied back to Rubisco, and the chloroplastic redox state and energy level are maintained. It is concluded that the optimal performance of photosynthesis is achieved via the provision of continuous CO2 supply to Rubisco by carbonic anhydrases and photorespiration.