With all the variety of models used for calculation of primary production of phytoplankton (PP) from remote sensing data, a choice of the most realistic one for a particular ocean region remains a non-trivial issue. In this work, we estimate PP in the Northeast Atlantic Ocean (200-510 N and 100-400 W) from 1998 to 2005 using three frequently used models: VGPM (Vertically Generalized Production Model), PSM (Platt and Sathyendranath Model) and Aph-PP model (Absorption Based Model). The modeled results are then compared with in situ observations of PP. The results show a close similarity in PP patterns obtained by different models, but the absolute modeled values differ substantially. In the Northeast Atlantic, PSM is found reproducing better the observed seasonal and spatial variability of PP as compared to the two other models. However, PSM slightly underestimates the PP values.