Image registration is a very important procedure in medical imaging analysis. However, the intensive computations involved in image registration have to some extent made it impractical for interactive use as well as limiting its general availability. This paper presents our current Grid project to facilitate image registration tasks. We have set up an image registration Grid by combining the attractive features of both Globus and Condor distributed computing environments. In order to make it more convenient to use, we have also developed a web interface for potential clients to specify and submit their image registration jobs to the Grid. The initial experiments in 3D breast MR images have shown encouraging results and demonstrated the suitability of Grid technology to this type of computationally intensive applications. The image registration Grid makes it much more straightforward for different institutes to use the identical registration program and protocols to register images consistently, quickly and efficiently. This can greatly improve data sharing and comparative studies in multi-centre trials. The Grid presented here could be an important step for clinical applications of image registration. Future work will focus on refining the Grid with the aim of upgrading it to a Grid Service and testing the system more extensively with medical imaging dataset.Image registration is an important procedure in medical imaging analysis. The purpose of image registration is to align one image (source image) to another (target or reference image) so that the possible misalignments between them can be minimized or eliminated, thus establishing spatial correspondence. Proper registration enables us to have a better understanding of the features of interest and integration of useful information. Applications are wide: ranging from image-guided surgery to atlas construction, segmentation propagation, monitoring changes over time and dynamic sequence analysis. The general procedure of fully automated image registration typically requires optimization of a function of the similarity between the target and source images. A large variety of image registration methods have been proposed for medical applications. In practice, image registration can be further classified as rigid and non-rigid according to the underlying transformation model. The former registers images by assuming that images are misaligned by translations and rotations only while the latter allows many more degrees of freedom. Compared to rigid registration, non-rigid registration can theoretically model more complicated deformations but requires more computation [1,2]. Rigid registration is normally used to provide a starting estimate before non-rigid registration in order to reduce computational time. Even so non-rigid registration still requires massive computational time, e.g. an average computation time up to 5 hours has been reported to register 3-D MR images of a single breast with a typical region of