Although DSC has been used extensively in the literature to measure the degree of cure and reaction kinetics, relatively little work has been performed to assess the limitations and accuracy of the DSC principle itself. The purpose of this paper is to examine two inherent limitations of DSC; control of sample temperature and measurement of heat flow. These limitations are a result of (1) the low thermal conductivity of the polymer sample, (2) heat transfer resistance at the interface between the pan and the sample holder, and (3) time lags inherent in the DSC control mechanism. These three contributions were examined in experiment and theory for a power compensated DSC with several types of sample pans. Procedures are presented for determining (a) the time lag of the DSC control mechanism, (b) the heat flow limit for a 1°C mean temperature change for a given sample and pan geometry, (c) an estimate of the mean temperature increase for a given exothermic heat flow, and (d) a correction for time lags in both the sample and DSC instrumentation. Dynamic DSC runs are shown to involve large sample temperature increases which may distort the calorimetric data. Inaccuracies due to large heat flow can be avoided by using either lower temperature or smaller samples.