Variable predictive model based class discrimination (VPMCD) is a pattern recognition method that utilizes the inner relations among characteristic values extracted from the original data. In this paper, VPMCD and independent component analysis (ICA) are combined with the correlation coefficient in order to diagnosis the rolling bearing fault. First, the ICA is used to analyze vibration signals with different fault categories, and the independent components are extracted from each category. Second, the correlation coefficients are extracted from the samples and independent components of each category. The sum of the absolute values of the correlation coefficients is used as a characteristic value. Finally, the VPMCD classifier is used to recognize and classify the faults. The experimental results show that this method can be effectively applied to rolling bearing fault diagnosis.