The measurement of diffusive gels in thin films (DGT) has recently been developed to assess metal bioavailability in soils. The DGT-method is based on diffusion in a porous matrix. To test the predictive capabilities of the method with regard to metal bioavailability, a study was set up with 28 soils having a variety of textures and amounts of zinc salts added. Correlation and regression analyses were performed to compare DGT-extracted zinc levels to zinc concentrations obtained by extraction with 0.01 M CaCl(2) and 0.43 M HNO(3), digestion with aqua regia and the zinc concentration in pore water. The amount of zinc extracted with CaCl(2) correlated well with DGT-extracted zinc levels in all soils spiked with different amounts of ZnCl(2). A similar correlation was not found for zinc concentrations in soil samples collected in the field. Experiments were performed to compare zinc content in organisms and in soils. The organisms tested were plants (grass, lettuce and lupine) and the hard bodied soil dwelling isopod Oniscus asellus. Good correlations were found between zinc accumulation in grass and lettuce and the C(E) (effective concentration) measured by a DGT-device, CaCl(2) extracted zinc and the zinc content in the pore water of all soils. The correlation with C(E) was not significant for lupine, neither for spiked soils, nor for field soils (p< or =0.001). Zinc levels in the isopods were not significantly related to any set of zinc measurements. From a synthesis of all results obtained it is concluded that the DGT-methodology does not have an additional value in predicting bioavailability of zinc in terrestrial ecosystems as compared to conventional extraction methods.