Dilatometry test systems are commonly used for characterizing the transformation behavior in steels and induction heating is commonly the heating source. In these systems, the steel test article is assumed to have a uniform temperature throughout the sample. This is a good assumption for slow heating rates with small samples, however, for induction hardening cycles this may or may not be accurate. Using computer models, it is possible to predict the temperature dynamics of the sample, both radially and axially, during the thermal processing cycle (heating and cooling). O1 tool steel was utilized to characterize and model heating and cooling temperature gradients. Specimens instrumented with multiple thermocouples were induction heated and gas quenched. The test data and geometry were evaluated with 1- D and 2-D models to characterize transient temperature gradients. The goal of the modeling is to better characterize temperature corrections required when rapid heating and cooling processes are used to determine transformation behavior in induction hardenable steels.