The correct characterization of petroleum fractions is crucial to correctly describe the processes in which they are involved. Most of the properties required to complete the characterization of these mixtures can be determined by well established methods. However, in the case of molecular weight, it is frequently determined by empirical correlations because of the scarcity of reliable methods. This work aims at evaluating the suitability of different techniques to calculate the molecular weight of petroleum mixtures. Cuts from three different crude oils within the whole range of boiling temperatures were analyzed by gel permeation chromatography (GPC) and high temperature gas chromatography (HT-GC). The former is a direct methodology to obtain the molecular weight. The latter, however, requires a process calculation as it yields the distillation curve, which must be converted into composition distribution. In this work, this calculation was performed by means of two continuous distribution models of different complexity. The molecular weight values obtained by GPC and HT-GC-based models were in good agreement, despite their different approach.