The objective of this study was set out to characterize the polycyclic aromatic hydrocarbon (PAH) content in the atmosphere of an oil furnace carbon black manufacturing plant located in southern Taiwan. A standard semi-volatile sampling train, the PS-1 sampler, was used to collect samples from eight areas, including the feedstock oil unloading, furnace, filtering/micro-pulverization, pelletizing, packaging, office/outside, office/inside, and boundary area, respectively. For each area, side-by-side static samples were collected simultaneously and a total of 16 samples were obtained. For each collected sample, the adsorbent-retained PAH content and the filter-retained PAH content were used directly to determine the concentrations of gaseous-phase PAHs and particle-bound PAHs, respectively. The gas chromatograph/mass spectrometer (GC/MS) technique was used for PAH analyses, and a total of 21 PAH species were determined. Results show the gaseous-phase PAHs accounted for only 69.2% of the total PAH content for samples collected from the packaging area, which was significantly lower than those samples collected from the rest of seven areas (ranging from 96.3 to 99.7%). The result is not so surprising since the packaging area had the highest dust concentration due to the releasing of carbon black dusts during the packaging process. In this study, we further examine the contribution of gaseous-phase PAHS to the total benzo[a]pyrene equivalent (BaPeq) content from the health-risk assessment view of point. It can be found the contribution of gaseous-phase PAHs to the total BaPeq content (63.1%) was quite comparable to the corresponding contribution to the total PAH content for samples collected from the packaging area. However, a different trend can be found for samples collected from the other seven areas, where the contributions of gaseous-phase PAHs to the total BaPeq content (ranging from 67.7 to 93.4%) were lower than the corresponding contributions to the total PAH content. The above results can be explained by PAH homologues that contained in both gaseous-phase and particle-bound PAHs. It was found the gaseous-phase PAHs contained higher fractions of less carcinogenic low molecular weight PAH homologues, whereas particle-bound PAHs contained higher fractions of more carcinogenic high molecular weight PAH homologues. Considering the contributions of gaseous-phase PAHs to both total PAH content and total BaPeq content were well above 50% for the eight studied areas, it is concluded that both particle-bound and gaseous-phase PAHs should be included for assessing the exposures of carbon black workers. (C) 2002 Elsevier Science B.V. All rights reserved.