Aims Evaluate crop rotation diversity, perenniality, carbon (C) inputs, and C input quality as predictors of soil organic carbon (SOC) concentration and aggregate mean weight diameter (MWD). Methods At a crop rotation trial in its 37th year in Ontario, Canada, species in rotations included corn (C, Zea mays L.), alfalfa (A, Medicago sativa L.), soybean (S, Glycine max (L.) Merr.), winter wheat (W, Triticum aestivum L.), and red clover (rc, Trifolium pratense L.). Rotations were: CC, CCAA, CCSS, CCSW, CCSWrc, AA. Soils (0-20 cm) were analyzed for aggregate MWD, aggregate C, and SOC concentrations. We estimated C inputs from historical yields and C input quality (C:N, lignin: N, or NMR-derived index) from structural plant tissues. C stabilization efficiency was estimated as the ratio of SOC stock per unit total or root C input. Results Crop rotation diversity failed to increase SOC concentrations or aggregate MWD. Perennialized rotations (CCSWrc, CCAA, AA) maintained the numerically highest SOC concentrations, and root C input increased SOC concentration. Out of 12 statistical tests relating C input quality to C stabilization efficiency, only 3 indicated a positive effect and 6 tests indicated a negative effect. Conclusions Including a perennial forage such as alfalfa and limiting soybean frequency promotes aggregate MWD and SOC concentration more so than optimizing crop rotation diversity. Quality of structural plant inputs does not explain differences in C stabilization efficiency, possibly due to overriding influence of living root inputs.