To evaluate the adequacy of the equilibrium partitioning concept in predicting metal bioaccumulation, a soil invertebrate species was exposed in 20 Dutch field soils with moderate metal contamination. Earthworms (Eisenia andrei) were kept in the soils for 3 weeks under laboratory conditions. Bioconcentration factors (BCFs) for six metals (Zn, Cu, Pb, Cd, Cr, Ni) and for As were calculated as the ratio of body- and solid-phase metal concentrations. Multivariate statistical analyses suggested that the BCFs for As, Cd, Cu, and Zn are governed by the same soil characteristics that determine equilibrium partition coefficients between the soil solid phase and the pore water. This suggests that uptake of metals is either direct from the pore water or indirect through an uptake route closely related to pore water. Regression equations were derived for predicting BCF values as a function of easily determinable soil characteristics. By means of internal validation it was shown that the equations obtained can be used for predictive purposes within the range of soil properties encountered in the dataset. Due to a lack of data, external validation was possible only in a qualitative sense.