Manganese(Mn) deficiencies are common in soils on the Loess Plateau of China. This research provided essential information on improving Mn availability in semiarid soils through agricultural practices. Twelve cropping system and fertilization treatments were designed in a 28-year experiment. The cropping systems included long-term fallow, continuous winter wheat cropping, pea(1 year)-winter wheat(2 years)-millet(1 year) rotation(crop-legume rotation) cropping, and continuous alfalfa cropping. The fertilizer treatments under the cropping systems included no-fertilizer control(CK), application of P fertilizer(P), application of N and P fertilizers(NP), and application of N and P fertilizers and manure(NPM), but the NP treatment was excluded in the continuous alfalfa cropping system. Available Mn and Mn fractions of soil samples(0–20 and 20–40 cm depths) were measured and further analyzed quantitatively using path analyses. Results showed that the crop-legume rotation and continuous alfalfa cropping systems significantly increased available Mn compared with the fallow soil. Compared with the no-fertilizer control, manure application increased available Mn in soil of the continuous wheat cropping system. Across all treatments, the averaged content of mineral-, oxide-, carbonateand organic matter-bound and exchangeable Mn accounted for 42.08%, 38.59%, 10.05%, 4.59%, and 0.09% of the total Mn in soil,respectively. Cropping significantly increased exchangeable Mn in soil and the highest increase was 185.7% in the continuous wheat cropping system at 0–20 cm depth, compared with the fallow soil. Fertilization generally increased exchangeable and carbonate-bound Mn in soil. Carbonate-bound Mn was the main and direct source of available Mn in soil, followed by exchangeable and organic matterbound Mn. These results indicated that crop-legume rotation cropping, continuous alfalfa cropping and application of manure, have the potential to promote Mn availability in soils of rainfed farmlands. This research provided essential information on improving Mn availability in semiarid soils through agricultural practices. Twelve cropping systems and fertilization treatments were designed in a 28-year experiment. The cropping systems included long-term fallow, continuous winter wheat cropping, pea (1 year) -winter wheat (2 years) -millet (1 year) rotation (crop-legume rotation) cropping, and continuous alfalfa cropping. The fertilizer treatments under the cropping systems included no-fertilizer control (CK), application of P fertilizer (P), application of N and P fertilizers (NP), and application of N and P fertilizers and manure (NPM), but the NP treatment was excluded in the continuous alfalfa Cropping system. Available Mn and Mn fractions of soil samples (0-20 and 20-40 cm depths) were measured and further analyzed quantitatively using path analyzes. Results showed that that crop-legume rotation and continuous a Compared with the no-fertilizer control, manure application increased available Mn in soil of the continuous wheat cropping system. Across all treatments, the averaged content of mineral-, oxide-, carbonateand organic matter-bound and exchangeable Mn accounted for 42.08%, 38.59%, 10.05%, 4.59%, and 0.09% of the total Mn in soil, respectively. Cropping significantly increased exchangeable Mn in soil and the highest increase was 185.7% in the continuous Carbonate-bound Mn was the main and direct source of available Mn in soil, followed by exchangeable and organic matterbound Mn. These results indicated that crop-legume rotation cropping, continuous alfalfa cropping and application of manure, have the potential to promote Mn availability in soils of rainfed farmlands.