This paper analyses the impact of a public research and development (R&D) subsidy to private firms in Mexico. My estimates suggest that the subsidy has a positive impact on personnel allocated to innovation activities, but it does not have an effect on stronger measures of R&D performance, such as research personnel, patents or private R&D spending. I argue that awarded firms would have performed their planned R&D projects in case they were not granted the public funds. Additional public funds seem to be invested to allocate more personnel on already planned projects, but not to carry out additional ones. Specifically, I analyse the Programa de Estimulos a la Innovacion (PEI) subsidy. The program's rules set a grade threshold below which no R&D projects get the grants and above which some projects are granted. This granting process allows to use a fuzzy regression discontinuity approach to identify causal inference.