ALOS-2 PALSAR-2 data acquired in full polarimetric mode at the rice heading and well-grown seasons were analyzed to classify agricultural areas according to their use. Eigenvalue-eigenvector and four-component decompositions were used to classify the PALSAR-2 data and discriminate different agricultural parcels. Threshold analysis was applied for the polarimetric decomposition components. Polarimetric decomposition parameters, alpha angle, double bounce scattering component ratio, and surface scattering component ratio, were applied to classify areas according to their use for paddy rice or other crops including soybean. The same threshold value used for analysis using the double bounce scattering component ratio was useful for all PALSAR-2 data analyzed in this study. However, the efficient threshold for alpha angle discrimination was different between the analyzed datasets. Differences in data observation direction, incidence angle, and the growth situation of paddy rice at observation time are conceivable reasons for this difference.