Objective To compare three simulated models（a commercially available model manufactured by Blue Phantom，a homemade gelatin model and a homemade tofu model） in the teaching of ultrasound-guided puncture technology，and the feasibility and teaching effect of the three models in the teaching programe were compared, so as to provide reference for the selection of appropriate teaching model. Methods 30 students from Nanjing medical university were enrolled into the study.All participants were recruited following an ultrasound guided regional nerve block tutorial.Following the tutorial，they attempted ultrasound guided nerve blocks using three different task trainers.Participants were assigned to each station in counting order but ultimately completed all three stations.After performing ultrasound guided nerve blocks using all three task trainers，the instructor evaluates the training effects，and the study participants were asked to give a numerical score for the following categories：realism，durability，cleanliness，image recognition degree and preference.Results The instructor evaluated the training effect according to the clarity of needle-track development in the process of students' operation, and there was no statistically significant difference among the three groups.The Blue Phantom model was favored in regard to realism，durability and cleanliness，respectively.There was no significant difference in image recognition scores among the three groups.Prior to revealing the cost of the task trainers，40% of the participants preferred the Blue Phantom model，while 40% preferred the homemade gelatin model and 20% preferred the homemade tofu model.After revealing the cost of each trainer，73% preferred the gelatin model.Conclusion The ultrasound-guided simulated puncture teaching needs to be evaluated according to the existing conditions of the unit，the actual cost and students' preferences，so as to select the most appropriate model.Among the puncture models that need to be designed independently, gelatin model has great advantages.