In this paper we develop a spatio-temporal registration algorithm for cardiac MR image sequences. The algorithm has the ability to correct any spatial misalignment between the images caused by global differences in the acquisition of the image sequences and by local shape differences. In addition it has the ability to correct temporal misalignment caused by differences in the length of the cardiac cycles and by differences in the dynamic properties of the hearts. The algorithm uses a 4D deformable transformation model which is separated into spatial and temporal components. The registration method was qualitatively evaluated by visual inspection and by measuring the overlap and surface distance of anatomical regions. The results demonstrate that a significant improvement in the alignment of the image sequences is achieved by the use of the deformable transformation model.