BACKGROUND: MULTIVARIATE ANALYSIS AS A CHEMOMETRIC TOOL HELPS GEOMETRICS FOR SOIL ANALYSIS IN PRODUCING INFORMATIVE RESULT. USING OF MODERN ANALYTICAL INSTRUMENT, CREATING LARGE AMOUNT OF DATA THAT BECAUSE OF SOME PROBLEMS SUCH AS STRONG VARIATION OF THE ELEMENTS CONCENTRATION IN ANALYZED SAMPLES AND MUTUAL INTERACTIONS HARDLY BE USEFUL WITHOUT PREPROCESSING [1]. EXPLORING SOURCES OF PRECIOUS METALS IS ONE OF THE MAIN CONCERNS OF GEOMETRICS. STATISTICAL ANALYSIS PROCEDURES CAN PROVIDE VALUABLE KNOWLEDGE FOR FINDING THE ORIGIN OF SUCH ELEMENTS AND INTERPRETATION OF INITIAL DATA [2].METHOD: IN THIS WORK, THE APPORTIONMENT OF AU IN ROCK SAMPLES BY STUDYING ON A COLLECTED DATA FROM SARILAR, AZARBAYEJAN SHARGHI, IS INVESTIGATED. IN ADDITION, THE DISTRIBUTION OF SOME PARAGENETIC ELEMENTS (CU, PB, ZN, AG, SN, MO, AS, SB, BI) IS DETERMINED.RESULT IN PRESENT STUDY, PRINCIPAL COMPONENT ANALYSIS (PCA) IS APPLIED ON A COLLECTION OF DATA CONSIST OF 130 SAMPLES AND LEVEL OF 10 ELEMENTS FOR ANALYSIS AND CLUSTERING. PCA RESULTS DEMONSTRATE A SPECIAL DISTRIBUTION PATTERN FOR SAMPLES. OBVIOUSLY, SAMPLES DIVIDED UPON THEIR GEOCHEMICAL CHARACTERISTICS AND CAN BE USEFUL IN FINDING PROSPECTING TARGETS SUCH AS AU.CONCLUSION: USING A PCA AS A SUPERVISED CLASSIFICATION METHOD FOR ANALYZING THE DATA SET PROVIDE AN INFORMATIVE RESULT OF A RAW GEOMETRICS DATA BY SUMMARIZING THE INFORMATION FROM A LARGE NUMBER OF COMPONENTS. UNFORTUNATELY, PRINCIPAL COMPONENT ARE OFTEN DIFFICULT TO INTERPRET. SO FOR PREDICTING MINERAL OCCURRENCE POTENTIALS AND IMPROVING ITS INTERPRETABILITY, IT WOULD BE BENEFICIAL TO APPLY OTHER CLASSIFICATION METHODS IN PARALLEL.