A description of the carbonate system has been incorporated in the MIRO biogeochemical model to investigate the contribution of diatom and Phaeocystis blooms to the seasonal dynamics of air-sea CO₂ exchanges in the Eastern Channel and Southern Bight of the North Sea, with focus on the eutrophied Belgian coastal waters. For this application, the model was implemented in a simplified three-box representation of the hydrodynamics with the open ocean boundary box ?Western English Channel? (WCH) and the ?French Coastal Zone? (FCZ) and ?Belgian Coastal Zone? (BCZ) boxes receiving carbon and nutrients from the rivers Seine and Scheldt, respectively. Results were obtained by running the model for the 1996?1999 period. The simulated partial pressures of CO₂ (pCO₂) were successfully compared with data recorded over the same period in the central BCZ at station 330 (51°26.05? N; 002°48.50? E). Budget calculations based on model simulations of carbon flow rates indicated for BCZ a low annual sink of atmospheric CO₂ (?0.17 mol C m^-2 y^-1). On the opposite, surface water pCO₂ in WCH was estimated to be at annual equilibrium with respect to atmospheric CO₂. The relative contribution of biological, chemical and physical processes to the modelled seasonal variability of pCO₂ in BCZ was further explored by running model scenarios with separate closures of biological activities and/or river inputs of carbon. The suppression of biological processes reversed direction of the CO₂ flux in BCZ that became, on an annual scale, a significant source for atmospheric CO₂ (+0.53 mol C m^-2 y^-1). Overall biological activity had a stronger influence on the modelled seasonal cycle of pCO₂ than temperature. Especially Phaeocystis colonies which growth in spring were associated with an important sink of atmospheric CO₂ that counteracted the temperature-driven increase of pCO₂ at this period of the year. However, river inputs of organic and inorganic carbon were shown to increase the surface water pCO₂ and hence the emission of CO₂ to the atmosphere. Same calculations conducted in WCH, showed that temperature was the main factor controlling the seasonal pCO₂ cycle in these open ocean waters. The effect of interannual variations of fresh water discharge (and related nutrient and carbon inputs), temperature and wind speed was further explored by running scenarios with forcing typical of two contrasted years (1996 and 1999). Based on these simulations, the model predicts significant variations in the intensity and direction of the annual air-sea CO₂ flux.