The examined components in the volatile phase of cigarette smoke increase with the puff number, in the same way as does the smoke condensate. The variation progresses similarly whether sampling with the Grob syringe or the smoking machine. However, when using the syringe we obtain lower values.The amount of particulate matter on a Cambridge filter has little or no effect on the retention of the volatile phase components by this particulate matter, at least after the smoking of the first cigarette.The volatile phase, resulting from the total phase collected in the syringe, gives, with the exception of benzene and toluene, higher values than those obtained by direct sampling through a Cambridge filter, both in the complete cigarette smoke test and in the puff-by-puff test.By sampling the different puffs each time through a new Cambridge filter we observe that values are systematically lower than those obtained by aspiration of the respective puffs through one and the same Cambridge filter.