Dissenting from existing multi-industries work, this paper particularly takes the pharmaceutical entities in the Chinese industrial enterprise database as a sample and uses the threshold regression to examine the impact of government R&D subsidies on private R&D investment, patent quantity and new product sales. We find that government R&D subsidies can effectively crowd in private investment. However, due to the crowd-out threshold is too high, the crowd-out effect does not actually exist in the pharmaceutical industry. A similar story happens to new product sales; but for the number of patents, there is only promoting effect.