The present paper investigates the effectiveness of R&D subsidies given to start-ups. Taking an aggregate view rather than evaluating a single program, we estimate the impact of R&D subsidies on start-ups' employment growth and their patent output. A unique data set on start-ups in the East German county of Thuringia allows us to focus on those start-ups that conduct R&D within the first three business years. We conduct propensity score matching to address the selection bias between subsidized and non-subsidized start-ups. We find that R&D subsidies lead to an increase in employment growth of about 66%. Furthermore, subsidized start-ups show a 2.8 times higher patent output. These estimates provide evidence for the additionality of R&D subsidies within the first three business years. Moreover, our analysis points to the special group of academic spin-offs which excels in the novelty of business ideas and patent activity. For some of these high-tech start-ups, no non-subsidized counterparts can be found. This might be attributed to the policy focus on academic spin-offs, which has led to a successful targeting of R&D support schemes.