The testing and evaluation program leading to the certification of a selected batch of polystyrene as ICTA Certified Reference Material GM 754 is described. Defined points from the glass transition curve were obtained, first in a preliminary program, then in 24 laboratories with the use of ten kinds of apparatus. Data were obtained in four heating rate ranges, 4-6, 8-12, 16-24, and 30-50°C/min. The initial departure was rejected for certification because of its relative lack of reproducibility and the comparatively subjective nature of its determination. The two defined points, Tband Tc, had unweighted mean values of 104.4 and 107.5°C, respectively, with standard deviations of 1.5 and 1.7°C, respectively. Data from individual investigators had a mean standard deviation of 0.53 ± 0.28°C. Separation of the data by heating rate, sample holder configuration, temperature sensor location and combinations of these disclosed that the reproducibility under a given set of conditions warranted certification; that the apparent heating rate dependence of the measured point (sensor location) from the sample and the inherent time dependence of the glass transition lead to an intrinsic contribution to the apparent heating rate dependence.