Purpose: Until recently, keyboard has been used as the primary input method for machinery operation system. But in recent years, numerous methods related to direct input interface have been developed. One of them is to measure the surface electric potential that generates on the skin surface during muscle contraction. Based of this fact, hand finger operation can also be recognized with the help of the surface muscle electric potential. The purpose of this study is to identify the hand finger operation using surface electromyogram (SEMG) during crookedness state of the finger. Design/methodology/approach: Two electrodes (Ag-AgCl electrode) were sticked randomly on the forearm muscles and the intensity of EMG signals at different muscles were measured for each crooked finger. Then depending on the intensity of the obtained electric potentials, a position was located and considered to have participated most actively during the crookedness state of that finger. Thus five locations on the forearm muscles were identified for five different fingers. Moreover, four different types of crookedness states were considered for each finger. Findings: In this experimental study, the electric current that generates on the skin during muscle activity was measured for different hand finger operations. As a result, it is found that there is a specified position related the maximum intensity of EMG signals for each finger. Practical implications: This paper cleared that the amount of crookedness of each finger can also be recognized with the help of surface EMG. It could be used as a machine interface technology in the field of welfare equipments, robot hand operation, virtual reality, etc. Originality/value: The objective of this research project was to develop the method of recognizing the hand finger operation and their crookedness states from surface electromyogram (SEMG).