In this paper, an adaptive output following control based on support vector machine (SVM) is proposed for a networked cooling process. Firstly, an approximated linear model of the cooling process is built through some thermal laws, where some parameters of the system are varied during operation. The parameters are identified by using SVM method. Based on the identified parameters, an adaptive output following controller with command generator tracker (CGT) is designed. By using the presented method, system stability and tracking performance can be ensured. Furthermore, experimental results are given to show the effectiveness of the proposed scheme.