This paper proposes a theoretical model of spatial duopoly, where the location, on the one hand, and the absorptive capacity of rms as function of their internal R&D investment, on the other hand, endogenously determine the maximum level of knowledge spillovers rms might absorb. Our goal is to test whether this new modelling of spillovers a ects the traditional outcomes in terms of rms location choices . We solve a three-stage game, where rms choose their geo- graphical location, prior to their level of internal R&D expenditures, and nally compete in prices. We found that, at the optimum, rms choose the same level of internal R&D and set the same price, inde- pendently of their location. Moreover, rms locate symmetrically and tends to agglomerate in the center of the market as the transportation costs increase, knowledge spillovers being in that case maximum.