Given the fact that fracture is a primary distress causing pavement failure, it is important to identify and characterize the fracture/cracking properties of asphalt concrete mixtures and to include them in pavement design processes. This study examined the testing variables for a reliable and practical semicircular bending (SCB) fracture test to evaluate the fracture characteristics of asphalt concrete mixtures at intermediate service temperatures. An integrated experimental-statistical approach was employed to identify testing variables by which repeatable SCB test results can be achieved. Using a typical Nebraska asphalt mixture, five critical testing variables (i.e., the number of testing specimens, specimen thickness, notch length, loading rate, and testing temperature) of the SCB test were investigated due to their significant effects on mixture fracture characteristics. Statistical analysis of test results indicated that approximately six specimens/replicates were a reasonable sample size that could properly represent asphalt concrete fracture behavior of a typical dense-graded mixture. Then, the coefficient of variation (COV) of the mixture fracture energy for six specimens was used to evaluate the effects of other remaining test variables. A range of a specimen thickness of 40 to 60 mm, a notch length from 5 to 40 mm, and a testing temperature between 15 and 40 C showed the reasonably low COV value of fracture energy at around or less than 10 %. The loading rates (0.1 to 10 mm/min.) attempted in this study did not show any significant differences in the testing repeatability.