Based on the microdata of 253 crowd innovation spaces (CISs) in Zhejiang province, this paper uses a continuous distance method to measure the level of spatial agglomeration, and uses zero-inflated negative binomial regression to analyze the influence mechanism. The study finds that the geographical distribution of the CISs in cities is significantly different, and the CISs are mainly distributed in the municipal districts of Hangzhou and Ningbo. The regression results show that the regional innovation capacity and the ecological environment both promote the geographic agglomeration of the CISs, but there is a substitution effect between them. Although it cannot directly promote the agglomeration of CISs, there is a significant complementary effect between policy guidance and innovation capacity. Accordingly, this paper puts forward policy suggestions from the aspects of nurturing innovation capacity and optimizing innovation incentives.