The migration of cells that are involved in the early (inflammation) stage of wound healing, such as neutrophils, macrophages, and T lymphocytes has been extensively studied for several decades. However, the migration of slower migrating cells, such as fibroblasts and endothelial cells, which are involved in the later stages of wound healing (proliferation and remodeling), is more difficult to study experimentally. In part, this is because such cells typically migrate one to three orders of magnitude more slowly than lymphocytes and neutrophils. Most models of cell migration are based on persistent random walks. A new technique for fitting cell migration data to persistent random walk models is presented, which accounts for variation within the population of cells and uncertainty in accurately determining the cells' positions. This technique enhances our ability to accurately determine such parameters as cell speed and effective diffusion coefficient for a population of cells, which cannot be directly measured experimentally. This model was developed to describe haptotaxis and haptokinesis of fibroblasts migrating on covalent peptide gradient surfaces. This technique enables quantitative distinctions to be made among different populations of fibroblasts migrating along different portions of peptide gradients, and responses to different peptides and different gradient magnitudes. By enabling accurate modeling of how these experimental parameters affect cell migration, this technique provides information on which engineers can base design of biologically activated surfaces for tissue engineering and wound healing applications.