Social practice teaching in colleges and universities combines theory and practice, transforms professional theory into practical behavior, and innovates students’ way of thinking, which can promote the teaching of colleges and universities as well as the cultivation of target talents. The big data era is innovating the traditional social practice teaching mode, and new opportunities are available for the development of social practice teaching. By analyzing the development form of social practice in colleges and universities in the era of big data, we propose to use the fuzzy C-mean clustering algorithm to deal with students’ social practice behaviors, to classify the type of students’ participation in social practice in colleges and universities, and to grasp the development status of the reform of social practice teaching in colleges and universities in the era of big data. The rating of dimensions of the social practice process in colleges and universities is used to determine the overall development level of social practice reform in colleges and universities. It also analyzes the differences in the process of carrying out social practice in colleges and universities, with the types of colleges and universities and categories of specialties as independent variables. When K is equal to 4, the contour coefficient is the largest, and the clustering effect is the best, so the students’ social practice participation can be reflected in four types: “need to improve”, “backward”, “hard-working” and “active”, combined with the overall score of social practice teaching reform in colleges and universities in the era of big data is 3.81±0.82 points, and the scores of all dimensions are above 3.75, and the overall development of social practice teaching reform in colleges and universities proposed in this paper is better.