Dermal absorption of contaminants from soils at federal contaminated sites in Canada was investigated using one hydrophile, (14)C-ethylene glycol (EG), and one lipophile, (14)C-nonylphenol (NP). In vitro dermal absorption of EG and NP was examined in dermatomed (0.4-0.5 mm) human skin using Bronaugh Teflon flow-through cells with Hanks HEPES buffered (pH 7.4) receiver solution with 4% bovine serum albumin (BSA). Tests were conducted under occlusive conditions with and without a commercial gardening soil spiked with EG or NP applied to skin at a soil load of 5 mg/cm(2). With percent absorption in skin depot included, a total of 9.9 + or - 6.28% (n = 6) and 34.8 + or - 8.47% (n = 6) absorption of EG with and without soil, respectively, and 20.6 + or - 5.56% (n = 7) and 41.1 + or - 6.46% (n = 7) of NP, with and without soil, respectively, were obtained. For tests without soil a reverse pattern was observed with significantly lower percent absorption into the receiver than depot with the lipophile NP, but significantly higher percent absorption in receiver versus depot for the hydrophile EG. This pattern was different in tests with soil, and caution needs to be exercised when extrapolating data from in vitro tests conducted without soil in human health risk assessments at contaminated sites.