Current standards for evaluation of asphalt mixtures for low-temperature performance are based heavily on asphalt binder properties. These standards ignore the interactions between the binder and the aggregate, both chemically and physically. These interactions have been shown to have significant influence on the cracking resistance of asphalt mixtures. This paper investigates three laboratory tests that evaluate cracking resistance of asphalt mixtures at low temperatures. The first test is the traditional tensile strength test, the indirect tensile test (IDT). The two other tests are fracture tests: semicircular bending (SCB) and disk-shaped compact tension [DC(T)]. Field cores were obtained from 10 pavement sections in Minnesota and Illinois for testing in the lab. Results from the three laboratory tests were compared and correlated with field performance data. Results showed that the tensile strength from the IDT did not vary significantly for the 10 mixtures. The SCB and DC(T) correlated relatively well with each other depending on the test temperature. The SCB showed the best correlation and highest Spearman's rank correlation coefficient when compared with the observed quantity of cracking in the field. It was concluded that the laboratory fracture tests, SCB and DC(T), were better suited for qualitative cracking performance predictions at low temperatures than the IDT. For quantitative predictions, application of advanced analytical or numerical models (or both) that take into account material properties, environmental conditions, and loading conditions is recommended. Also recommended is that additional studies comparing laboratory fracture parameters with field performance be conducted.