Butylate [S-ethyl bis(2-methylpropyl)carbamothioate] persistence was compared in three soils that were treated annually with four thiocarbamate herbicides in 1981 and 1982. Butylate degraded faster in soils treated two consecutive years with butylate + dichlormid (2,2-dichloro-N,N-di-2-prophenylacetamide) than in soils treated for 2 yr with EPTC (S-ethyl dipropyl carbamothioate) + dichlormid, cycloate (S-ethyl cyclohexylethylcarbamothioate), vernolate (S-propyl dipropylcarbamothioate) + dichlormid, or from previously untreated soils. Butylate + dichlormid applications in 1981 and 1982 enhanced butylate degradation in 1983 at all locations and reduced weed control at two locations. Previously untreated soil or soil previously treated with two annual applications of butylate + dichlormid was treated with 6 ppmw butylate, incubated for 14 days, analyzed for residual butylate, and bioassayed for herbicide activity. Bioassay indicated higher herbicidal activity than was predicted by the residual butylate concentration. A herbicidally active metabolite may be produced during the soil degradation of butylate.