PAHs are major constituents of creosote, and it has been suggested that their volatilization from treated wood could make a major contribution to the U.K. atmospheric emission inventories. This paper reports a study to elucidate the volatilization characteristics of selected PAHs from creosote-treated wood and to estimate the potential significance of this source to the U.K. atmospheric source inventory. Desorption kinetics from untreated and treated wood were measured at 4 and 30 °C for five PAHs (acenaphthene, fluorene, phenanthrene, anthracene, fluoranthene) using chambers in the laboratory. Rates of desorption followed first-order kinetics for all compounds and were higher at 30 than 4 °C. Mean ∑PAH fluxes varied from 2.57 ± 1.52 to 29.5 ± 6.1 mg/m2 treated wood/day at 4 and 30 °C, respectively. A prolonged study of the kinetics at 4 °C showed that the volatilization rate remained constant over about 7 weeks, after which >85% of the compounds still remained sorbed on the wood. Initial desorption rates were governed by partitioning between the substrate (wood) and air and subsequently by the rates of compound diffusion from the interstices of the wood. A strong negative correlation was observed when the percent increment in the desorption rate from 4 to 30 °C was plotted as a function of the octanol−air partition coefficient (KOA) (r 2 = −0.93); vapor pressure was strongly positively cor related (r 2 = 0.92). Desorption half-lives ranged from 0.7 to 31 yr at 4 °C and from 0.3 to 1 yr at 30 °C for fluoranthene and acenaphthene, respectively. Estimates of annual U.K. emissions for the compounds specified from freshly treated wood (i.e., ignoring long-term releases from previously treated wood) were subject to a number of important assumptions, but were in the range of 100 t ∑PAH. This is less than the annual U.K. emissions of ∑PAHs estimated for domestic heating (600 t) and believed to be similar to that from vehicle emissions (80 t) at the present time.