In this paper, we use a conditional-convergence econometric model to investigate whether the Cohesion Policy and the structural funds this policy mobilises, affect the European economies in such a way that the poorer regions catch up with the rich ones. In this model, regional convergence depends on policy treatment and regional economic structure, proxied by investment per capita and the demographic growth rate. The main originality of the model is its specification, dealing with temporal and spatial issues at the same time. Econometric estimations rely on a dataset of 143 EU14-NUTS1/NUTS2 regions observed over more than 25 years (from 1980 to 2005). Generalized Method of Moment estimation enables us to obtain consistent estimates of the beta-parameter along with estimates of the impact of regional policies and regional economic structure on regional growth. Our results suggest that Objective 1 programmes have a direct effect on regional GDP p.c. growth rates, whereas total structural funds do not. We interpret this result as an Objective 1 programme added-value, compared to total structural funds. However, these results do not mean that the non-Objective 1 structural funds have no impact on overall growth in the EU (e.g. through a technology diffusion effect) but they do not allow additional growth specifically in these regions, when we consider the spatial dependences. Moreover, consideration of the spatial dimension of the panel brings to light a still significant, but less important, impact of structural funds on convergence.