A decision making computer aided diagnostic (CAD) system has been proposed in this paper for the classification of focal kidney lesions using various wavelet energy descriptors .The focal kidney lesions are categorised into two kidney classes namely primary benign i.e. angiomyolipoma (AML) and primary malignant i.e. renal cell carcinoma (RCC). The study is performed on 47 kidney ultrasound images with 22 AML lesions and 25 RCC lesions. The multi-resolution wavelet based texture descriptors are calculated from area of interests (AOIs) of variable sizes by using different types of wavelet filters such as Haar, Daubechies, biorthogonal, symlets and coiflets filters. It has been seen that by using SVM classifier, Daubechies (db4) and symlets (sym3) wavelet energy descriptors gives the best overall classification accuracy of 82.6 %, with the individual class accuracy (ICA) values of 63.6 % for Angiomyolipoma and 100 % for Renal cell carcinoma.