Does science and technology evaluation (STE) policy reform improve universities' innovation performance? Research typically argues that more upstream R&D investment leads to more downstream innovation performance, but it is less clear what the trade-offs over time might be for targeted investments in universities influenced by STE reform. In 2013, China proposed STE reform measures focused on ‘innovation quality, efficiency, and contribution’. Using a difference-in-differences research design on a comprehensive longitudinal database of 62 universities spanning from 2009 to 2016, we show that this STE policy positively affects university basic research outputs and quality but weakens applied research outputs. This effect is pronounced in pilot universities with better resources. Empirical evidence suggests that the STE policy works mainly through mobilising the enthusiasm of human capital, improving R&D intensity in science and technology funds, and promoting industrialisation-oriented R&D projects. We consider the following possible perspectives on the mechanisms of change: knowledge asset development, economic competition, and a socio-political process. This analysis leads to theoretical developments about how basic versus applied science produces outputs in the Chinese context. The paper also shows that the STE policy promotes technology transfer in universities.