Field, greenhouse, and laboratory studies were conducted to evaluate the performance of starch xanthide (SX), sludge polymer (SP), and coventional formulations (CF) of benefin [N-butyl-N-ethyl-2,6-dinitro-4-(trifluoromethyl)benzenamine], oxadiazon {3-[2,4-dichloro-5-(1-methylethoxy)phenyl]-5-(1,1-dimethylethyl)-1,3,4-oxadiazol-2-(3H-one}, and prosulfalin {N-[[4-dipropylamino)-3,5-dinitrophenyl] sulfonyl]-5,5-dimethylsulfilimine} for the control of large crabgrass [Digitaria sanguinalis (L.) Scop. # DIGSA] in Kentucky bluegrass (Poa pratensis L.) turf. Turf injury was greatest with SP oxadiazon and prosulfalin formulations, while SX formulations of oxadiazon and prosulfalin caused decreased and /or delayed injury and provided control comparable to conventional formulations. Coarse SX granules containing prosulfalin caused less turf injury than fine granules, while the opposite effect sometimes occurred with SX oxadiazon. Differences in control were observed in the greenhouse when SX benefin formulations which varied in cross-linking agent and/or degree of substitution were compared to the conventional formulation on sandy and silt loam soils. Benefin SX formulations also demonstrated controlled-release properties, which improved large crabgrass control when compared to the conventional formulation in the greenhouse. This effect was short lived on silt loam but persisted on sand. SX granules cross-linked with Fe3+ extended benefin activity longer than H2O2 cross-linked materials on sandy soil only. Release of 14C-labeled benefin from SX matrices was altered by the extent of water imbibition, solvent characteristics, and granule size.