Ombrotrophic peat bogs are natural archives of atmospheric pollution, their depth profiles can be used to study the deposition chronology of harmful contaminants. Prerequisites for deriving historical deposition rates from the peat archive are that contaminants are persistent and immobile in the peat and that the applied dating technique is accurate. To examine these requirements and the accuracy of peat archives for polycyclic aromatic hydrocarbons (PAHs) 12 peat profiles were sampled in 4 bogs in Ontario, Canada, as well as surface peat in one bog. Additionally we carried out laboratory incubations; no degradation occurred over a 3-year period in these experiments. The standard deviations of PAN concentrations in surface samples and of PAH inventories in whole cores was approximately 30%, and concentrations in surface peat were on average 50% higher in hollows than in hummocks. No indications for mobility of PAHs were observed in peat. Temporal deposition trends inferred from peat cores were generally in agreement with trends derived from a sediment core sampled close by but deposition rates to the sediment were substantially higher. A major source of uncertainty was the rather coarse vertical sampling resolution of 5 cm which introduced substantial uncertainty in the dating of the individual segments. This caused variations of the deposition rates up to 70% per PAH between three replicate cores, and it also impedes the identification of deposition peaks. Overall, we conclude that peat cores are suitable archives for inferring atmospheric deposition trends, but due to their relatively low temporal resolution short-term events may not be identified and the development of sampling methods that allow a higher vertical resolution would greatly improve the performance of the method. The analysis of more than one core per site is suggested to provide a realistic estimate of the historic deposition and total inventories.