The Measures for Management of Soil Environment in Agricultural Land (Trial, Nov. 01, 2017, China) recently came into effect and highlighted the proper management of contaminated croplands to lower risks of exceedances of contaminants, especially toxic trace metals in agricultural produce. We aimed to develop local soil criteria for lead (Pb) in Hezhang county of southwestern China by the inverse use of reliable models linking Pb contamination levels between soils and vegetables. Dilute nitric acid (0.43 M) extraction, a new ISO standard (ISO-17586:2016) for extracting the geochemically reactive Pb fraction (Pb_NA), and calcium chloride (0.01 M) extraction (ISO-14255: 1998) for estimating the plant-available Pb (Pb_CC) were performed in fifty historically polluted and newly Pb-spiked soils with differing soil types, properties (pH 4.1–8.0), and total soil Pb levels (Pb_T, 20–6153 mg kg⁻¹). Greenhouse experiments for Brassica pekinensis L., and in-situ soil porewater measurement for Pb were conducted to investigate the mechanism of Pb uptake, and to establish reliable Pb soil-plant relationships. The results indicated that about 83% of the variation for Pb concentrations in vegetable (Pb_CL, 0.009–1.06 mg kg⁻¹) was contributable to free Pb²⁺ activity in soil porewater, which was mainly influenced by pH and dissolved organic matter. Pb_CL was satisfactorily predicted using Pb_NA and key soil properties (adj. R² 0.852). Soil Pb criteria for Pb_T and Pb_NA are then derived based on food standard. The full implementation of criteria derived for Pb_NA (i.e., 27–127 mg kg⁻¹, soil pH 5.5–8.0) can avoid the exceedance of Pb in 95% of cabbage samples in this study, 95% of cabbage cultivars by model extrapolation, and one widely cultivated root vegetable, radish, in the study region. We provide a successful case study that has effectively tackled the challenge for the complexity of the soil management in contaminated croplands.