Pseudo random number generators have many applications in different fields such as system simulation, communication and cryptography. Recent two decades chaos and chaotic maps receives more attention to use these maps for designing pseudo-random numbers generators. Chaotic maps have many interesting properties such as random behavior, ergodicity and sensitivity to initial conditions. In this thesis some generators based on chaotic maps have been evaluated and discussed and some advantages and disadvantages of the methods are obtained. Then we suggested a new pseudo random binary generator by using of logistic and 3D cat maps. To evaluate the suggested generator, the output of the generators have examined by statistical tests and due to the conclusions, we can say that the generators have proper and acceptable random behavior. Keywords: Pseudo random number generator, statistical tests, chaos, chaotic maps, logistic map, cat map