This paper attempts to check the existence of geographic and industry distance effects, alongside other microeconomic determinants, on firms’ decisions to engage in R&D collaboration. Physical distance is defined by geographical coordinates while the measure of industry distance is based on the trade intensity between sectors. The model specified here refers to the combined spatial autoregressive model with autoregressive disturbances and it is estimated through the spatial two stage least square procedure. The results show that both geographical and industry proximity, positively affect the decision to cooperate in R&D.