To solve the problem of fuzziness and randomness in fault diagnosis of tolerance analog circuits, a diagnosis method based on the multi-feature cloud correlation matrix is proposed. First, a combined cloud model is proposed according to the characteristics of analog circuits. Then, by constructing the cloud correlation matrix, the cloud membership degree of different circuit features is evaluated synthetically. Considering the different contributions of different features to the overall cloud correlation, the cloud correlation matrix is weighted according to the within class and between class scatter. Finally, the identification of the fault is realized by obtaining the cloudrelated sequences of the feature vectors to be identified and the failure modes. The example analysis shows that the proposed method has higher diagnostic efficiency for circuit soft faults under the tolerance condition.