In this study electromyography (EMG) signals obtained from index finger movements are analysed and classified in two separated methods and results are compared as a pre-study of dexterous anthropomorphic prosthesis hand project. The index finger differs from other hand digits for its tendon structure and higher mobility capabilities. The relation between the EMG signal and movement of the index finger is a key to understanding how the finger performs the tasks such as grasping and postures. Signals are measured from two different muscle groups and classified by grasping and positions. Finite State Machine (FSM) and Artificial Neural Network (ANN) are used for classification. Also the study indicates the effects of signals from index finger motions on a myoelectric controlled prosthetic hand.