Abstarct: Identification of vocal disorder has been a vital role in our life nowadays. Acoustic analysis can be useful tool to diagnose voice disorders as a complementary technique to other medicine methods such as Laryngoscopy and Stroboscopy. In this thesis, the aim is diagnosis of vocal fold disorder and type of disease using speech signal analysis. Types of the vocal disorders are: noduls, polyp, edma, paralysis, paresis, hyperfunction, reflux, erythema, polypoid degeneration and A-P squeezing. The proposed method has three stages which are feature extraction, feature reduction and classification. The first and second stages perform a critical role in performance and accuracy of the classification systems. The feature extraction procedure is implemented using the wavelet-packet decomposition of the coefficients of speech signal .The energy and entropy Shannon feature, obtained from the coefficients in the output nodes of the optimum wavelet packet tree, are used. Linear discriminant analysis (LDA) method is proposed for feature reduction stage . Support vector machine (SVM) is used as a classifier for evaluating the performance of the proposed method. The results show the priority of the proposed method in comparison with the current methods. Experimental result show that recognition accuracy is 100% for separating healthy from unhealthy class and recognition accuracy is 98.88% for separating disease from eath orther.