For numerical simulations of the injection moulding process, an accurate description of the specific volume is needed to predict differential shrinkage during and after moulding, which causes thermally induced stresses and controls the dimensional accuracy and long term dimensional stability. For amorphous polymers, for which often it can be assumed that cooling rate dependence can be ignored, standard techniques enable measurements of the specific volume as a function of temperature and pressure. For semi-crystalline polymers the situation is more complicated since the specific volume depends strongly on the degree of crystallinity which itself depends on the thermo-mechanical history, i.e. temperature and pressure (for quiescent crystallization). This requires the use of a combined experimental-numerical technique to interpret the data and to determine the specific volume. Standard equipment can only be used at relatively low cooling rates. Since high cooling rates are present during injection moulding, improved experimental techniques must be designed. A setup based on the confining fluid technique is build, which can reach cooling rates up to 60[Ks −1 ] and pressures up to 20 · 10 6 [P a]. During an experiment, the specific volume is measured together with the temperature history and pressure. Using an accurate model to calculate the crystalline structure, together with a specific volume model which depends on this structure, enables the determination of model parameters. Comparing both the measurements and the model predictions, leads to the conclusion that modeling of the crystallization kinetics, results in accurate predictions of the specific volume.