At present, there is no specific teaching method for doctor-patient communication for oncology residents. This study combined BOPPPS (bridge-in, learning objective, pretest, participatory learning, posttest, and summary) teaching model and SBAR (situation-background-assessment-recommendation) communication model to try a new teaching and assessment model of doctor-patient communication, aiming to explore and improve the teaching method of doctor-patient communication for oncology residents. Ninety residents were randomly divided into two groups: the experimental group (n = 45) was trained with the BOPPPS teaching model combined with the SBAR communication model, the routine teaching method was adopted in the control group (n = 45). Indicators of assessment included doctor-patient communication skills, satisfaction with teaching, and patient satisfaction. The scores for doctor-patient communication ability were significantly better in experimental group than control group (p < 0.05). The degree of satisfaction degree (very satisfied + satisfied) of the experimental group were higher than control group (p < 0.05). The overall teaching satisfaction of the experimental group was 100.00%, the control group was 77.78%, p < 0.001. Patients' satisfaction with the residents in the experimental group (93.3%) was significantly higher than control group (80.0%), p = 0.043. The application of the BOPPPS combined with the SBAR can effectively enhance residents' doctor-patient communication ability and improve teaching satisfaction and patient satisfaction. This new model can effectively improve resident physician‒patient communication ability, which is worth actively promoting in clinical teaching work.