The decoupling of crop and livestock in space is currently threatening the sustainable development of agriculture. One of the major challenges is to identify the imbalanced space between cropland and livestock farms. This paper proposes a new methodology to identify such areas at the raster scale, without being limited by traditional administrative division restrictions. The method takes into account nutrient flow, economic cost, and spatial distributions of cropland and livestock farms. The methodology was applied to the Heihe River Basin of China as an example. First, we calculated the nutrient demands on croplands and the supply of livestock manure. Second, we proposed a new method to capture the imbalanced space between cropland and livestock farms. The method was improved on the Gaussian two-step floating catchment area (2SFA) method. To our knowledge, this is the first time such a method has been applied in agriculture. Third, we applied the maximum coverage model to select sites for manure transfer hubs in the imbalanced space of cropland and livestock farms. In this way, the distance of transferring manure can be effectively reduced. Our results showed that the total amount of manure nutrients was sufficient to meet the requirements of local crops in over 90% of the counties located within the Basin. Despite the potential benefits of manure as a fertilizer, its adoption could be more expensive than chemical fertilizers on almost 92% of the croplands due to the high transportation costs associated with spatial separation. However, the identification of 119 transfer hub sites significantly enhanced the accessibility of manure for returning to croplands, ultimately resulting in the optimization of 32% of the previously imbalanced spatial relationship between cropland and livestock farms. The method facilitates raster-scale identification of imbalanced space between cropland and livestock farms, with wide applicability across regions globally.