This paper adopts the USEM Employability Model as a theoretical model to classify college students’ employability into disciplinary comprehension, comprehensive skills, self-efficacy, and metacognition, and to classify college mathematics teaching into basic theory courses, applied skill development, and professional expansion. The questionnaire data were collated, the reliability and validity of the questionnaire were examined, and the differences in university mathematics teaching as well as students’ employability were analyzed in four school types: double first-class colleges and universities, general undergraduate colleges and universities, specialized colleges and universities, and others. The correlation analysis of the dimensions of university mathematics teaching and employability is conducted, the correlation coefficient is calculated, the logistics regression analysis of university mathematics teaching and employability is launched, and the moderating effect analysis is conducted for the double first-class colleges and universities and the general undergraduate colleges and universities respectively. School type showed significance (p < 0.05) for college mathematics teaching overall and for college students’ employability overall and for the four dimensions of subject comprehension, comprehensive skills, self-efficacy and metacognition. The explanatory degree of the demographic characteristics variable in the overall regression analysis on college students’ employability is 33.5%, value is 0.335. After adding the college mathematics teaching variable, is 0.672, which gives us that college mathematics teaching positively affects employability. That is to say, improving university mathematics teaching can positively promote students’ employability, and the effect is more significant in double first-class colleges and universities.
