Emotional forecasting was a psychological research hotspot in recent years. Bias not only existed in forecasting self emotions, but also in forecasting others’, which can lead to interpersonal empathy gap. Van Boven and Loewenstein presented the dual judgment model of emotional perspective taking, implying that self forecasting bias and similarity judgment deviation between themselves and others are the main reasons of interpersonal empathy gap. Based on the dual judgment model, this study established an interpersonal emotional forecasting model, which can not only explain the formation of interpersonal empathy gap, but also help to overcome interpersonal empathy gap. Taking the typical interpersonal forecasting phenomenon (teachers forecasting students’ emotions) as example and selecting typical educational situations, this research explored the formation of interpersonal empathy gap as well as ways to overcome interpersonal empathy gap to verify the interpersonal emotional forecasting model. Study 1 (85 teacher participants, 364 students participants) selected caring students and delaying class as forecasting situations. Study 2 (37 teachers and 37 students) selected positive expectation as forecasting situation, where the teachers were asked to forecast students’ emotional responses. The results indicated that whether they are forecasting groups or individual student, there exists a significant correlation between teacher’s self-forecasting and forecasting the student’s emotions. As the empathy anchor, self-forecasting directly affected teachers’ forecasting on the students' emotions. While the influence of empathy anchor on forecasting accuracy has two-sidedness, the extremeness of self-forecasting would lead to a still greater empathy gap between teachers and students. Study 3 (61 teachers, 68 students) selected preparing for tests as forecasting situation, while the teacher’s empathy strategy was manipulated by instructions. It indicated that empathy strategies significantly influenced the teachers' emotional forecasting accuracy. Teachers thinking only from their own perspective and neglecting to imagine emotions from the student’s perspectives would induce greater forecasting bias. Otherwise, teachers placing themselves on students’ position would overcome egocentric bias, thereby effectively closing the empathy gap. Effective emotional prediction strategy was directly presented in this research with realistic effects and sustained the development of the dual judgment model of emotional perspective taking. Interpersonal emotional forecasting model was conducted and verified. This study laid a theoretical basis for further revealing the characteristics and mechanisms of interpersonal emotional forecasting process.