This paper raises the question of the relationship between innovation and productivity conditional to ICT use and R&D activities during the innovation process. Concretely, we wonder whether R&D activities and ICT use considered as innovation inputs play a role in determining the probability of introducing technological and non technological innovations and whether these innovation outputs conditional to ICT use and R&D lead to a higher level of labour productivity. In this adaptation of CDM model extended with ICT use, we consider a four equations model that relates labour productivity to innovation outputs, innovation outputs to R&D and ICT use, and R&D and ICT use to their determinants. Our robust three-step estimations underline the benefits of taking into account the joint endogeneity of the key variables of the whole system. Unlike the previous empirical literature, we introduce in the model a large set of indicators of ICT use to capture the degree of variety and sophistication of ICT and distinguish two types of R&D activities, internal and external R&D instead of constituting aggregated measures of ICT use and R&D. We observe that while confirming the acknowledged ‘innovation-enabler’ role of some ICT, the results point out the fact that not all increases in ICT investments translate into equivalent increase in firm capacity to introduce new products/processes or in an improvement of innovative performance. In addition, the productivity effect of product innovation appears only at the moment of its commercial success, measured as the turnover percentage of sales generated from new or improved products. Furthermore, estimation results confirm the acknowledged belief that new or improved organizational arrangements, conditional to ICT platforms, lead to a subsequent improvement of product quality, timeliness, reduce waste, transactions and coordination costs, which could, in turn, result in an improvement of the labour productivity.