Objective To develop an intelligent classification system for diabetic foot ulcer based on Wagner grades, and to test its consistency with the assessment result by wound therapist. Methods According to 1 042 manually labeled foot ulcer images, deep learning method was used to identify and segment the areas of foot ulcer images, and an intelligent classification system was deve-loped to realize automatic grading of foot ulcer. Totally 150 foot ulcer images of 68 hospitalized diabetes patients were selected as the verification group, and the consistency of Wagner grading results of the images by the intelligent classification system and international wound therapist was evaluated. Results The weighted Kappa value of diabetic foot intelligent classification system and wound therapist was 0.643 (P<0.01). Conclusion The intelligent classification system of diabetic foot ulcer based on Wagner grades can realize clinical homogeneous evaluation, assist clinical medical staff to assess and grade diabetic foot ulcer, then implement targeted treatment and nursing, so as to improve nursing quality and effect.