In this paper, we use an inverse methodology to parameterize a box model of the Red Sea to study carbon exchanges and transport. In this application, we use simultaneously the information contained in physical and biogeochemical tracer profiles such as temperature, salinity, dissolved inorganic carbon, alkalinity, oxygen, and phosphate. Since the Red Sea is subject to a seasonal monsoon regime and because we are interested in resolving exchanges from the surface to the bottom, we have decoupled the temporal regime into two seasons : winter (November to May) and summer (June to October). We present the two solutions which include advective and mixing (eddy diffusion) exchanges as well as new primary production, decomposition, carbonate formation, and dissolution. The circulation is reversed from one season to the next. The net biological exchanges (new production) of carbon, compatible with the dynamical solutions, are almost null for the organic processes (negative in summer, positive in winter); the budgets for the mineral processes are negative for both seasons. This result is consistent with the view that carbonates play an important role in the budget of carbon in Red Sea. Finally, taking advantage of richer data sets for the summer season, we present two box designs for this period. It appears that transport solutions are sensitive to the choice of box topology, whereas biochemical exchanges are roughly the same for both designs.