<p>High nitrogen (N) inputs in croplands and their runoff into waterways have led to proliferation of the harmful algal blooms in coastal zones. Wetland restoration is regarded as a “nature-based solution” for N reduction. To answer the question “where should we restore wetlands to maximize N removal”, we used wetland datasets, digital elevation model data, N input data, and literature knowledge of wetland N removal at wetland and sub-basin scales to identify gaps and priorities for wetland restoration in the mid North China Plain. The results showed that 1) the spatial mismatch between cropland N surplus and wetland N removal limits the contribution of wetlands to water quality. The total N removal by wetlands in the study area is approximately 376.86 kt N/yr, or close to 13% of the cropland N surplus for the study area; 2) wetland position supplies key information for identifying gaps in wetland N removal that wetland area alone can’t capture; 3) wetland complexes at sub-basin scale rather than wetland area should be considered as the management target. With the same increase (1%) in wetland area, targeted restoration scenario provides a 2.89% increase in wetland N removal, which is 54 times more N removal than that of random placement scenario. These results underline that combining supply–demand budgets analysis with the perspective of wetland complexes is an important step in identifying restoration gaps. The framework proposed in this study is spatially explicit, effective, repeatable, and thus has potential to improve the wetland restoration strategies in China.</p>