Image Registration is one of the important techniques of digital image processing and its goal ;s the correspondence of two images for the same scene taken on two different times such that any pixel that has the same coordinates in both images wm always indicate the same position in the scene. One of the important applications of image registration is in the field of medical X-ray images where its aim is to discover the changes which has occurred on the scene better an the first and the second image, There are many classical techniques which are used for this purpose that require the identification of some points in one of the images and the corresponding points in the other image (tie points). The purpose of this is to see how useful it is to use genetic programming image registration where it is left for the system to determine automatically tie points (control points) within some specified region. In this thesis, genetic algorithms were studied and developed using many different algorithms including, simple crossover, double crossover, fuzzy crossover, mutation, inversion, and also mixed techniques of the above algorithms were implemented. In addition, fuzzy logic was used in the initialization process. Two genetic registration methods are proposed and implemented. A software package called Digital Image Registration System (DIRS) was designed to implement the image registration methods for correcting 2-dimensionaI gray-level images using genetic algorithms. Also, DIRS was supplemented by two Classical methods: the manual method and the correlation method. These methods were implemented for comparison purposes. Results from the use of the suggested methods and. Results from the Classical methods were compared, and encouraging results are obtained. The minimum requirements which have to BE satisfied to run DIRS are an IBM-AT or anything compatible, a VGA monitor, and a mouse.