The rock components are very important for the logging interpretation of complex lithology reservoir. In order to obtain the ratio of rock matrix components point by point quickly and accurately, a fuzzy clustering method was discussed by taking the Biyang dolostone reservoir as an example. First, logging parameters were processed by principal component analysis and the purpose is to reduce the number of dimensions. Samples were processed by routine standardization, mean processing standardization, and logarithmic transformation standardization. The cumulative variance contribution rates of the first three principal factors are 86.07%, 96.97%, 96.71%, respectively. Then the samples are clustered and the classification number of k-means clustering is discussed. Finally, the lithologic component is point by point and quantitatively obtained by using the idea of fuzzy mathematics and calculating the degree of membership. The results show that the method is effective and suitable for the analysis of a large number of well logging data. It is worth to point out that this method is also effective when the new materials of logging technique are lack of.