Permeation of tecnazene into and through human cadaver skin in vitro was assessed using a CC-MS method employing HS-SPME for receptor solution analyses. Two doses of tecnazene dissolved in acetone, corresponding to 103 and 864 microg/cm2 of tecnazene, were applied to skin mounted on Franz diffusion cells and placed in a fume hood. Cells were either occluded with aluminum foil or left unoccluded. Total absorption of tecnazene (dermis + receptor fluid) after 48 h was 2.2-6.1% of the applied dose for the unoccluded treatments and 22-33% for the occluded treatments. Potentially absorbed dose including all tecnazene that may have eventually permeated the skin ranged from 10% unoccluded to 42-53% occluded. Accumulation in the receptor solutions was satisfactorily described by a working diffusion model after upward adjustment of the partition coefficient for tecnazene in all skin layers by a factor of 5-16 versus a priori values. However, residual amounts of tecnazene in both the epidermis and dermis were higher than those estimated from the model, suggesting the existence of tissue binding not accounted for in the calculation. The results indicate that the diffusion model as presently calibrated may significantly underestimate both systemic absorption and skin concentrations of highly lipophilic compounds, as predicted from data generated from in vitro skin permeation assays. Model predictions could be improved by better accounting for partitioning into the epidermis and dermis.