This paper is concerned with the determination at the number of branches in PVC by means of pyrolysis-gas chromatography from the comparison of the benzene and toluene peaks in gas chromatograms. It was theorized that if the toluene peak was related to the branches, because they were one carbon atom long, the toluene and benzene peaks would have been formed according to the same mechanism, i. e. , dehydrochlorination, polyene pi -electron delocalization, ring formation, and chain scissions. Slow pyrolysis experiments were carried out to see if the toluene and benzene peaks would appear in the same time and temperature. Thermogravimetric analysis (tga) in combination with gc was used to determine quantitatively the volatiles evolved during thermal degradation of rigid PVC samples. Results are presented and evaluated. It is concluded that the measurement of the toluene/benzene ratio of the products of flash pyrolysis on pure PVC cannot be safely related to the number of branches. A more recent paper by Liebman et al. shows that during the ″time-resolved″ pyrolysis gc of vinylchloride-propylene copolymers, toluene and benzene are produced simultaneously; then, the toluene/benzene ratio may be used as a measure of the copolymer composition.