Based on the data of 281 cities from 2005 to 2016, we constructed difference-in-difference (DID) models to study the impact of national innovative-city-pilot policy on urban innovation from the three dimensions of dynamic effect, heterogeneity, and influence path. The research found that: (1) The innovative-city-pilot policy has significantly improved the level of urban innovation. This conclusion is still valid after a series of robustness tests. (2) The policy does not have a lag effect, but it has a certain dynamic and continuous effect. (3) This policy is more conducive to innovation in provincial capitals or sub-provincial cities, and has little effect on innovation in other cities; there is no significant difference in the policy effects on eastern, mid-western cities, but the average policy effect in eastern cities is slightly higher than in the mid-western cities. (4) The impact mechanism test shows that the pilot policy affects urban innovation through government fiscal expenditures, the degree of urban industrial agglomeration, and the level of human capital.