Under the background of environmental pollution and energy depletion, online teaching is the key means to reduce the waste of resources and achieve the sustainable goal of education. This paper constructs a fuzzy comprehensive evaluation model to scientifically estimate the satisfaction of online teaching mode and introduces a multiple linear regression model and structural equation model respectively to empirically analyze the factors affecting satisfaction. The results show that: (1) Online teaching in colleges and universities has achieved remarkable results, and students’ satisfaction is at a good level. According to the fuzzy comprehensive evaluation method, the comprehensive satisfaction scores for the dimensions of teacher questioning, flipped classroom, platform resources, student learning, and teacher engagement are in the range [80, 90]. (2) Teaching platform resources, student discussion and online skills are the key factors of online teaching satisfaction. According to the results of multiple linear regression analysis, the factors A11, B21, and C21 are significant at the 1% level in terms of significance level, with p-values of 0.008, 0.010, and 0.004, respectively. (3) The influence paths of teaching knowledge, teaching interaction and teaching platform on teaching satisfaction are different. According to the calculation results of structural equation model, teaching knowledge and teaching interaction have a positive impact on teaching satisfaction, and the regression coefficients for both are 0.123 and 0.930, respectively, while the regression coefficient of teaching platform on teaching satisfaction is -0.168, which has a negative impact. Finally, it is necessary for universities to build a three-dimensional element of a “teacher-student platform” to continuously optimize the path of online teaching satisfaction.