A study was conducted to determine whether test specimen produced using silicone (Si) rubber molds affect the Superpave low-temperature grading parameters, S(60) and m-value, determined using the bending beam rheometer (BBR; AASHTO-TP1). Six laboratories participated in this study. Four unmodified asphalt binders with low-temperature performance grades between -6 deg C and -30 deg C were tested. The thickness of each test specimen was measured using the "flip" technique. The results showed that the Si rubber molds produce significantly thinner beams than the required 6.35 +/- 0.05-mm thickness. This affects the calculated S(60) values but not the m-value. Aluminum (Al) molds were also used in this study according to AASHTO-TP1. It was found that the stiffness values S(60) calculated in the BBR test are significantly lower when Si rubber molds are used to produce test specimen as compared with those test beams produced using Al molds. The cause is the peculiarity of the calculation process specified in AASHTO-TP1. AASHTO-TP1 requires that the thickness of the test beam must always be assumed to be 6.35 mm for computing S(60). This study suggests that the test beams produced using the Si rubber molds are significantly thinner; therefore, assuming a thickness of 6.35 mm to calculate S(60) for these beams produces a negative bias. It is shown that when thickness is measured using the flip technique and the measured thickness is then used to calculate S(60), there is no statistically significant difference between the S(60) values of the test beams produced using either the Al or Si rubber molds.