Based on the runoff plots located in Yanggao County, Shanxi Province, the effects of different vegetation types, including bare land, wasteland, Caragana korshinskii, Hippophae rhamnoides, Pinus tabuliformis and Medicago sativa, on soil quality in loess hilly area in North Shanxi Province were assessed by principal component analysis. The results showed that eight soil physicochemical indicators (soil texture, porosity, capillary porosity, organic matter, total nitrogen, alkalihydrolysis nitrogen, available phosphorus and available potassium) could be represented by three common types of soil quality factors, which were organic matter factor, soil texture factor and pore factor. The scores of both the organic matter factor and pore factor under the four artificial vegetation types were all higher than in the wasteland. Except for M. sativa, soil texture factor scores under the artificial vegetation were also higher than in the wasteland. Scores of all three common quality indicators in the bare land were the lowest. Soil integrated assessment scores for M. sativa, C. korshinski, H. rhamnoides, P. tabuliformis, wasteland, and bare land were 0.596, 0.584, 0.495, 0.481, 0.305, and 0.194, respectively. This meant that the four types of artificial vegetation could improve soil quality and M. sativa and C. korshinskii were more effective than others. Bare land led to soil degradation.