This paper constructs a food logistics system evaluation model based on generalized regression neural network. It first establishes an evaluation index system for food logistics system and studies the standardization of relevant indexes; then it explores related theories of generalized regression neural network and establishes a food logistics system evaluation model based on generalized regression neural network; finally a fairly satisfying test result is acquired through numerical example examination. The test result shows: the evaluation model proves to be simple and practicable and is effective in evaluating food logistics system. This paper provides food producers with an effective tool to select evaluate and manage food logistics system.