A model of biological production in the euphotic zone of the North Atlantic has been developed by coupling a seven-compartment nitrogen-based ecosystem model with a three-dimensional seasonal general circulation model. The predicted seasonal cycles of phytoplankton, zooplankton, bacteria, nitrate, ammonium, primary production, and particle flux have been compared to data from Bermuda Station ''S'' and Ocean Weather Station ''India''. Bearing in mind the simplicity of the model and the paucity of data, the results are encouraging. However, deficiencies in the physical model lead to winter nitrate values at Bermuda being overestimated, and at both positions the predicted magnitude of the spring phytoplankton bloom was too high. Simulations were carried out with different detrital sinking rates and it was found that a sinking rate of 10 m d-1 gave the best agreement with observations. The model was used to investigate the factors affecting the population growth of phytoplankton and it was found that the model supported the generally held theory that the spring bloom is initiated by the cessation of physical mixing. After the bloom, phytoplankton are controlled by zooplankton grazing. At Ocean Weather Station ''India'' the model reproduced the observed high summer nitrate levels and suggested that these high values are caused by a combination of high vertical nitrate transport, ammonium inhibition of nitrate uptake, and zooplankton grazing control. The model demonstrated the critical importance of zooplankton in understanding ecosystem dynamics and highlights the need for more observational data on the seasonal cycles of zooplankton biomass and growth rates.