Models describing the partitioning of atmospheric oxidized mercury (Hg(II)) between the gas and fine particulate phases were developed as a function of temperature. The models were derived from regression analysis of the gas-particle partitioning parameters, defined by a partition coefficient (K p ) and Hg(II) fraction in fine particles (f PBM ) and temperature data from 10 North American sites. The generalized model, log(1/K p ) = 12.69-3485.30(1/T) (R 2 = 0.55; root-mean-square error (RMSE) of 1.06 m 3 /μg for K p ), predicted the observed average K p at 7 of the 10 sites. Discrepancies between the predicted and observed average K p were found at the sites impacted by large Hg sources because the model had not accounted for the different mercury speciation profile and aerosol compositions of different sources. Site-specific equations were also generated from average K p and f PBM corresponding to temperature interval data. The site-specific models were more accurate than the generalized K p model at predicting the observations at 9 of the 10 sites as indicated by RMSE of 0.22-0.5 m 3 /μg for K p and 0.03-0.08 for f PBM . Both models reproduced the observed monthly average values, except for a peak in Hg(II) partitioning observed during summer at two locations. Weak correlations between the site-specific model K p or f PBM and observations suggest the role of aerosol composition, aerosol water content, and relative humidity factors on Hg(II) partitioning. The use of local temperature data to parameterize Hg(II) partitioning in the proposed models potentially improves the estimation of mercury cycling in chemical transport models and elsewhere.