The metabolism of Amidosulfuron in soil under aerobic conditions was studied in three laboratory metabolism studies using six different soils. In the first study with four different soils (sandy loam, sand, loamy sand, silt loam) two metabolites accounting for > 10 % AR were detected. One of it (“metabolite B”) reached a maximum of 49.6 % AR after 7 days in a loamy sand soil and was identified as HOE 101630, which is formed by demethylation of the parent substance. Another metabolite (“metabolite A”) reached a maximum of 25.8 % AR after 100 days. This metabolite was later identified as HOE 128870, which is formed by hydroxylation from HOE 101630. Three other metabolites (metabolites “C”, “D” and “E”) were detected. With a few exceptions metabolites D and E could not be separated by HPLC analyses. Metabolite C and HOE 101630 could also not be separated at later sampling points during the course of the study. Attempts have been made to estimate single values for these metabolites by extrapolating the proportion of the single metabolites from those sampling points where separation was possible to the later sampling points. Unknown metabolite “C” occurred in maximum amounts of 7.7 % AR (day 14), 4.8 % AR (day 14), 4.3 % AR (day 3) and 6.2 % AR (day 3) in the four soils. Unknown metabolite “D” occurred in maximum amounts of 5.0 % AR (day 28) 8.1 % AR (day 21) and 8.8 % AR (day 70; extrapolated value) in three of the soils, respectively. In one of the soils this extrapolation was not possible for metabolite “D” as no single values at all for this metabolite were available in this soil. But from the data it can be assumed that in this soil the amount of metabolite “D” would be very small. Metabolite E occurred in maximum amounts of 8.6 % AR (day 49), 9.1 % AR (day 21) and 12.1 % AR (day 35; extrapolated value) in three soils, respectively. In the fourth soil no single value was available for this metabolite and thus no extrapolation was possible. From the data it can be assumed that also in this soil the amount of metabolite “E” is in the range of 8-12 %. Metabolite “E” was later identified as ring-hydroxylated amidosulfuron (later assigned as AE 1569309) by comparing chromatograms (relative retention times, HPLC conditions) from the soil study with chromatograms from studies on residues in wheat plants. It is an important fact that both, metabolite AE F128870 and AE 1569309, can not be synthesised and identification is only possible by indirect methods, as no standard solutions are available. Therefore separate studies on fate and behaviour or ecotoxicology are not possible with these metabolites. In the second study with one soil (sandy clay loam) metabolite HOE 101630 reached a maximum of only 5.2 % AR after 14 days. “Metabolite A” (AE F128870) reached a maximum of 16.6 % AR after 70 days. All other not identified metabolites accounted for not more than 2.1 % AR. In the third study with one soil (loamy sand) metabolite HOE 101630 reached a maximum of 8.4 % AR after 7 days, and two not identified metabolites “U2” and “U4” reached maxima of 38.6 % AR after 56 days and 10.8 % AR after 41 days, respectively. Due to its retention time metabolite “U2” was proposed by the notifier to be AE F128870. Comparing chromatograms and relative retention times metabolite U4 was found to be identical with metabolite “E” of the first study and thus can be identified as metabolite AE 1569309. One identified minor metabolite (AE F094206) reached a maximum of 1.5 % AR. Under aerobic conditions the formation of CO2 was in the range of 3 to 47% after 91-100 days. Not extractable radioactivity amounted for up to 16 - 59% after 91-100 days.