Abstract It is important to know how tobacco additives behave when cigarettes are smoked, whether they transfer intact to the smoke or whether there is any decomposition during smoking. Pyrolysis-GC-MS is a technique that can be focussed upon the effects of combustion from a single material free from interference from the complex mixture of different components present in the smoke. However, because pyrolysis is a model technique, the results need to be validated by comparison with cigarette smoke chemistry. In a previous paper we presented such a method for modelling the smoke chemistry from a burning cigarette using pyrolysis-GC-MS. The transfer and the extent of degradation of anisole, p-anisaldehyde, benzaldehyde, isoamylisovalerate, methyl trans-cinnamate and vanillin within a burning cigarette were estimated using this pyrolysis method. When these data were compared with results from smoke studies from 14C-analogues of the materials, the high levels of transfer predicted by pyrolysis were found to be generally consistent with the smoke chemistry data. However, there were still two outstanding issues. Firstly, there was some ambiguity in the labelled study about whether vanillin actually transferred without degradation or not. Furthermore, the results from the 14C-labelled study showed a greater extent of degradation for p-anisaldehyde than that indicated from the pyrolysis experiments. The purpose of the current study was to present some new information obtained to address these questions by better understanding the effect upon the smoke chemistry from adding vanillin and p-anisaldehyde, and the relationship between the smoke chemistry and the pyrolysis results. Components were identified in the smoke from cigarettes loaded with p-anisaldehyde and vanillin labelled with 18O and 13C. The extent of degradation from each additive was estimated by identifying labelled degradation products in the smoke. Because there was a clear distinction between the mass spectra from the labelled compounds and their natural counterparts, the major degradation products from the labelled compounds could be readily identified in the chromatograms of the corresponding smoke extracts. Evidence of some degradation at above the transfer temperature was indicated for both additives. The amount of degradation was found to be less than 1% for p-anisaldehyde and only 0.1% for vanillin. This low level of degradation was acceptably consistent with the intact transfer values of 97% and 100%, respectively, obtained by pyrolysis.