By regulatory authorities the rat is considered to be a suitable animal model to predict the percutaneous absorption of hazardous substances in humans. In our study, the percutaneous penetration of 2-butoxyethanol (BE) and toluene was compared in different rat models. Intradermal microdialysis and static diffusion cells were used in in vivo and in vitro experiments with haired Wistar and hairless Lewis rats. Microdialysis experiments showed a steady-state penetration for BE and a penetration maximum for toluene in both rat strains at approximately 60 min after beginning of exposure. However, in diffusion cell experiments the penetration of the test compounds in both rat strains increased until the end of exposure (4 h). Additionally, in microdialysis experiments BE penetrated in hairless rats in a higher amount than in haired rats (factor: 1.4; P < 0.01), for toluene it was just the opposite (factor: 1.9; P < 0.001). In diffusion cell experiments, the penetrated amounts of both compounds were higher in hairless rats compared to haired rats. The fluxes for BE were in diffusion cell experiments at a factor of 14.5 (haired rat) and 18.1 (hairless rat) higher than in microdialysis experiments, the difference factor for toluene was 2.6 (haired rat) and 12.9 (hairless rat). The lag times indicate a significantly faster penetration in microdialysis experiments compared with diffusion cell experiments (P < 0.001). There are great differences in percutaneous penetration behaviour between the techniques and the rat strains. The diffusion cell method has difficulties to describe the percutaneous penetration kinetics, whereas microdialysis describes it more reliable. Due to these differences the reliability of a conversion factor for the transfer of percutaneous absorption data from rat to human skin, as proposed in the literature, is questionable.