Abstract : In our proposed work we are predicting the gender from the palm image. Using ROI segmentation, we can extract the palm region, from the segmented palm region extract the geometrical features and calculate various parameters. from the various parameters will train the machine and classify the trained data using SVM. finally generating the confusion matrix to assess the efficiency. Here we are proposing a method which extract the feature of the segmented palm region, considering it as a square and calculate the sides and diagonal, based on that we will classify the palm based on gender. © BEIESP.