The perennial rhizomatous weed Cirsium arvense (L.) Scop. is difficult to manage on semiarid organic farms. Our objective was to quantify the impact of eight 4‐year crop sequences crossed with standard‐ and reduced‐tillage on C. arvense occurrence (presence/absence), stem density, and aboveground biomass at two semiarid sites in Montana. The sequences represented a range of crop competition from high (multiple years of perennial forage [alfalfa, Medicago sativa L.]) to low (2 years of continuous fallow), with intermediate sequences consisting of different annual species. Final‐year spring wheat (Triticum aestivum L. emend. Thell.) was planted in all sequences to determine impacts on subsequent cash crop production. Through time, alfalfa and double fallow sequences reduced C. arvense density and biomass where it was established, although its invasion into new areas increased in the double fallow. Final‐year C. arvense occurrence and stem density were lower in the alfalfa sequence compared with six and four annual crop sequences, respectively (p < 0.05). Final year C. arvense biomass was higher at one site than the other, although not in the double fallow sequence. Wheat grain yields differed in response to crop sequences at the two sites: at one site, grain yield was lowest in the alfalfa sequence, especially when standard‐tillage was used, while yields were highest in the alfalfa and double fallow sequences at the other site. Using perennial forages in semiarid organic systems can be effective for managing C. arvense, but subsequent cash crop yield may be depressed.