We conducted a series of laboratory microcosm incubations with [(14)C]-labeled 17beta-estradiol (E2), estrone (E1) and 17alpha-ethinylestradiol (EE2) in 17 different natural soils to characterize hormone mineralization. A significantly higher mineralization was observed for E1 (2.0-37.6%) and E2 (4.2-50.2%) than for EE2 (0.5-2.6%) in all test soils after 21 days. Soil physical or chemical parameters were not related to estrogen mineralization. Although sorption parameters varied greatly for E2 (K(F)=21.9-317.5 mL g(-1)), for E1 (K(F)=46.0-517.5 mL g(-1)) and for EE2 (K(F)=29.9-326.1 mL g(-1)) this apparently did not control estrogen bioavailability since it showed no effects on hormone mineralization. In order to elucidate the controlling factors, experiments with combined additions of radiolabeled estrogens and different substrates were conducted. Additions of ammonium nitrate or alanine to soil samples generally increased EE2 mineralization, thus indicating N-limitation. Additions of glucose induced higher E2 and EE2 degradation in comparison to control samples which is attributed to co-metabolism. Additions of saw dust, catechol or streptomycin influenced the microbial population in the test soils and affected the mineralization of E2 and EE2. Thus, we clearly demonstrate that different microbial communities are responsible for E2 and EE2 degradation in soils. We suggest that EE2 is mineralized by white-rot fungi and E2 by bacteria.