It is well known that cracking resistance of hot-mix asphalt (HMA) mixtures is influenced by many factors, such as the asphalt binder type, binder content, aggregate gradation, aggregate type, and volumetric parameters of mixtures. The Texas Transportation Institute's overlay tester (OT) was used to evaluate cracking performance in terms of number of cycles to failure (N_OT), and an experimental design was undertaken to cover the many combinations of aggregates, gradations, and asphalt binders. Through extensive laboratory tests and statistical analyses, the most significant factors that affected cracking performance were identified. Corresponding N_OT prediction models were developed with high R² values (approximately .9). With these models, the minimum asphalt content for different HMA mixtures was estimated, and the lower limits of asphalt content during mix design were recommended. Finally, the models and recommended values were preliminarily verified with asphalt overlay test sections on Interstate 20 near Atlanta, Texas.