Abstarct: In this thesis, neural network weights using imperialist competitive algorithm optimized and for finger-vein recognition has been used. Neural network computational tool for applications such as pattern recognition and prediction and...Is and is in spired by the human nervous system. Imperialist competitive algorithm is one of optimization algorithms introduced in 2007 by Mr.Ismail Atshpz Gargari and Caro Lucas and from the social phenomenon of competition between imperialist countries and countries under their domination is in spired. From optimization algorithms in neural network for improving weights and the network topology can be used. In this thesis, for improving weights has been used. In recent years the need for people identification systems baxsed on biometric features has been a large increase. These systems replace traditional identification systems in modern society and in business, office and household...,been used. Biometric identification systems compared to traditional systems have high security. One of the people identification systems, which has along history and is most common biometric system is fingerprint. Biometric systems baxsed on human characteristics, behavior and physiology work. Physiological characteristics such as fingerprint have high accuracy from behavioral features such as voice. In some cases, the combination of multiple methods of biometric identification are used. Advances of biometric technology are increasingly and other biometric characteristics human is discovered. The thesis of imperialist competitive algorithm, improving the weights of the neural network is used for the recognition of the finger vein.The imperialist competitive algorithm by performing a series of steps, the initial weights of the neural network and the network start to work. In this thesis, from two databaxses PKU and USH-FVP, have been used.