We performed laboratory incubation experiments on the degradation of gaseous phase carbonyl sulfide (OCS) by soil bacteria to determine its sulfur isotopic fractionation constants ((34)ε). Incubation experiments were conducted using strains belonging to the genera Mycobacterium, Williamsia, and Cupriavidus isolated from natural soil environments. The (34)ε values determined were -3.67 ± 0.33‰, -3.99 ± 0.19‰, -3.57 ± 0.22‰, and -3.56 ± 0.23‰ for Mycobacterium spp. strains THI401, THI402, THI404, and THI405; -3.74 ± 0.29‰ for Williamsia sp. strain THI410; and -2.09 ± 0.07‰ and -2.38 ± 0.35‰ for Cupriavidus spp. strains THI414 and THI415. Although OCS degradation rates divided by cell numbers (cell-specific activity) were different among strains of the same genus, the (34)ε values for same genus showed no significant differences. Even though the numbers of bacterial species examined were limited, our results suggest that (34)ε values for OCS bacterial degradation depend not on cell-specific activities, but on genus-level biological differences, suggesting that (34)ε values are dependent on enzymatic and/or membrane properties. Taking our (34)ε values as representative for bacterial OCS degradation, the expected atmospheric changes in δ(34)S values of OCS range from 0.5‰ to 0.9‰, based on previously reported decreases in OCS concentrations at Mt. Fuji, Japan. Consequently, tropospheric observation of δ(34)S values for OCS coupled with (34)ε values for OCS bacterial degradation can potentially be used to investigate soil as an OCS sink.