Objective　To understand the effect of standardized training for resident doctors and to improve its evaluation method. Methods　A questionnaire survey was made among 89 resident doctors including 46 directors and 46 matrons. Kirkpatrick Model was used to analyze the question-naires in four dimensions including reaction layer， learning layer， behavior layer and achievement layer. Data of reaction layer were analyzed by ANOVA and data of learning layer and behavior layer were compared by paired t-test. Results　Reaction layer： the overall satisfaction of the resident doc-tor for training is about 4.09. Learning layer： There were significantly statistical differences in the test scores before and after training（P=0.00）. Behavior layer： There were significantly statistical differences in resident doctor’s ability and quality before and after training（P=0.00）. Achievement layer： both the complaint rate and the accident rate of the 89 resident doctors were 0 while all of them passed the practicing physician exam. Conclusions　Standardized training can improve the quality and ability of resident doctors. Kirkpatrick evaluation model can reflect the effect of the standardized training for resident and is a good tool to analyze the effectiveness of standardized training for resident.