Abstract: Based on the related data of listed companies in China's smart cities from 2011 to 2015, by means of panel threshold regression technique, this paper makes an empirical study of R&D investment's heterogeneous threshold effect on firm performance. The results show that:(1)R&D investment has an obvious nonlinear inverted U-type influence on firm performance, and only modest R&D input could promote firm performance, while if the investment is excessive, it could generate an inhibitory effect;(2) Constrained by enterprise scale, R&D input's effect on firm performance is featured by nonlinear U-style. Thus, it can be seen that only when the enterprise scale goes above a certain threshold, R&D investment can advance firm achievement;(3)Under the threshold of capital structure and the threshold of capital density, R&D input has a significant inverted N-type and nonlinear N-style effect on firm performance. At the same time, the optimal capital structure and capital intensity level range could be identified;(4) Under the ‘New Norm’, to further activate R&D input's positive effect on firm performance, we must grasp these key points-to pay equal attention to both R&D and management capability, to further strengthen and expand enterprise development, and to keep the modest capital structure and higher capital density level.