The aim of the Programme Océan Multidisciplinaire Méso Echelle (POMME) was to study the formation and subduction of 11°–13°C waters in the northeast Atlantic (21°–15°W and 38°–45°N). An extensive oceanic and atmospheric data set was collected over 1 year during the period September 2000–October 2001. Owing to the importance of energy and water exchanges between the top layers of the ocean and the atmosphere in the subduction process, a surface heat, freshwater, and momentum budget has been computed combining the use of satellite products, in situ data, and atmospheric model outputs. This data set has been compared and validated with observations collected from a moored buoy and an instrumented mast onboard a research vessel. Each component of the net heat, freshwater, and momentum flux has been individually evaluated, and turbulent fluxes were computed with a state-of-the-art bulk flux algorithm deduced from turbulence measurements made during the experiment. We have adopted a 5 km grid spacing to take into account the oceanic mesoscale variability. The annual domain-averaged heat flux is positive (+33 W m^-2), indicating a heating of the ocean, whereas model estimates (European Centre for Medium-Range Weather Forecasts (ECMWF) and the French operational weather forecast model, ARPEGE) indicate a negative (cooling) budget (−9 W m^-2 and −25 W m^-2, respectively). Sensitivity tests of the parameterization used and the sea surface temperature used place the accuracy of the budget to about 10 W m^-2 The freshwater budget is negative, implying a freshening of the ocean, as in the ECMWF model. Our assessment proves that sea surface temperature patterns condition the mesoscale patterns of the heat budget, a feature that is not reproduced by models.