Purpose- The purpose of this paper is to investigate the determinants of R&D investment at the level of the firm. Design/methodology/approach- A balanced panel of 215 Italian manufacturing firms over the 1995‐2000 period has been used to test the technology‐push, the demand‐pull and the endogenous skill‐bias hypotheses. Econometrically, both the GMM‐SYS estimator and the Least Squares Dummy Variable Corrected (LSDVC) estimator (a recently proposed panel data technique particularly suitable for small samples) have been used. Findings- Results support the well‐established technology‐push and demand‐pull hypotheses and, furthermore, supply evidence for the role of skill endowment in increasing a firm’s R&D investments. Research limitations/implications- A limitation of the study concerns the measure of skills which is here, as in previous economic literature, simply the ratio between productive (blue‐collar) and non‐productive (white‐collar) workers. Another limitation of this contribution concerns its limited generalisability: data come from relatively large Italian manufacturing firms, i.e. the service sector and SMEs are not considered. Practical implications- Consistently with the related managerial and economic literature, the basic result is that current skill endowment may significantly and positively influence a firm’s current R&D decision; therefore, adequate education and training policies may indirectly induce an increase in corporate R&D investment. In terms of managerial implications, this means that HRM may be seen as an indirect strategy for improving a firm’s R&D effort and ultimately for improving its performance through innovation. Originality/value- While there is a well‐established literature investigating the so‐called Skill Biased Technological Change, few microeconomic empirical studies have been devoted to test the reverse relationship. The paper aims to fill this gap, testing whether higher skills may induce higher R&D expenditures.