Résumé : Light and dark rates of net protein synthesis by Phaeocystis pouchetii colonies were measured in Belgian coastal waters during the spring bloom of 1984 by following I4C-bicarbonate and l5N03/13NH4: assimilation into cellular proteins during natural 1ight:dark cycles. A close agreement was found between the 2 tracer methods. Ambient inorganic nitrogen was shown to be the most important factor controlling protein synthesis in the light, which was considered to proceed at a constant rate in the whole euphotic layer. Protein synthesis in the dark, on the other hand, was found to be controlled by both inorganic N and 'previous light history' of the cells. The relation between specific protein synthesis and external inorganic nitrogen obeyed Michaelis-Menten kinetics for both light and dark processes. The Michaelis-Menten constant K, was 4 µmoles N 1-1 while the maximum specific protein synthesis rates p,, were respectively 3 and 1.8 % h-' for the light and dark processes. The relation between protein assimilation in the dark and 'previous light history' could be expressed by a linear relation between dark specific protein synthesis and the number of hours spent at 3 8 J cm-'h-', the energy threshold for the occurrence of dark protein synthesis. From the knowledge of these control mechanisms a simple mathematical model was established in order to calculate daily integrated net primary production from the hourly variations of light intensity in the water column and ambient inorganic N concentrations. These estimates have been compared with those calculated from classical methods for measurements of primary production.