The crystallization of a high-density polyethylene was analyzed with differential scanning calorimetry (DSC) measurements. An intense transcrystallinity was observed at the contact between the polymer and the DSC pans. The modification of the crystallization kinetics induced by this phenomenon was studied as a function of cooling rate and sample thickness. We point out that most of the theoretical predictions of our previous model could be checked. The crystallization temperature was a function of the sample thickness and could be also correlated with the thickness of the transcrystalline zones. The shapes of the DSC traces were complex and correlated with the amount of trancrystallization. The usual interpretations of such DSC curves were not accurate. We conclude that specific experimental procedures must be proposed to understand and correctly use such measurements.