Dynamic infrared imaging (DIRI) is an emerging technology for the early detection of breast cancer. In this method time-series of thermal breast images are obtained. The patient motion in the time-series can distort the DIRI analysis in such a way that the detection of breast cancer becomes impossible. Image registration can be used to eliminate the patient motion from the time-series data. In this thesis, two different registration algorithms were tested: Thirion's demons algorithm and an algorithm based on an affine transformation. Furthermore, a combined method where the affine method is used as a pre-registration step for the demons method was tested. The algorithms were implemented with Matlab and their performance in the task of registering a time-series of thermal breast images was evaluated using four different performance metrics. The registration algorithms were implemented for time-series data of 20 healthy (no malignant lesions) subjects. The demons method outperformed the affine method and is recommended as a suitable tool for time-series registration of thermal breast images. The combined method achieved slightly improved results compared to the demons method but with significantly increased computation time.