A study using several types of sampling systems was conducted in the chamber facility at the Pierce Laboratory to compare the determination of nicotine in environmental tobacco smoke generated by volunteer smokers. The sampling systems used included filter packs, annular denuders, sorbent beds and passive samplers. Total nicotine determined using the various sampling systems was generally in good agreement. Agreement among samplers was also generally good for gas phase nicotine. The most notable exception was the determination of nicotine using a nonpassivated stainless steel passive sampler where the results were low due to adsorption of nicotine by the sampler. Agreement, but with poor precision, was seen for the determination of particulate phase nicotine using a Tenax microtube sorbent sampling system and two different annular denuder systems. However, loss of particulate nicotine to the gas phase occurred during sampling with the filter pack systems, and determination of particulate phase nicotine by these systems was in error. Because greater than 95% of the nicotine was in the gas phase, this loss of particulate nicotine did not significantly affect the determination of gas phase nicotine using a filter pack sampling system. (Copyright (c) 1990 by the American Chemical Society.)