To construct a soil health evaluation index system based on the minimum data set，comprehensively evaluate the impact of different tillage methods combined with straw returning to the field on soil health，screen out the conservation tillage model suitable for the northeast black soil region，a long-term tillage experiment was conducted in Gongzhuling city，Jilin province. Three tillage methods were adopted，namely rotary tillage and straw not returning，no tillage with straw returning，and moldboard plowing with straw returning. Perennial native grassland was selected as the control of natural and healthy soil. In 2019，soil samples of 0-10 cm in the surface layer were collected to determine the physical，chemical and biological indicators of the soil. Multivariate statistical methods such as principal component analysis，factor analysis，and partial least squares discriminant analysis(PLS-DA)were adopted to screen key indicators to construct the minimum data set. Combined with linear and nonlinear scoring functions，the soil health index(SHI)was calculated. The applicability of SHI was constructed by different methods in assessing soil health status in the northeast Black soil region，and the soil health status under different tillage methods was comprehensively evaluated. The results indicated that:(1)Compared with straw not returning，no tillage with straw returning significantly increased the soil carbon and nitrogen contents. Specifically，soil organic carbon and total nitrogen increased by 40.9% and 36.7%，respectively，compared to straw not returning(P<0.05)；meanwhile，the soil microbial biomass carbon and phosphorus acquisition enzyme activity in the no tillage with straw returning treatment also significantly increased，with increments of 182.0% and 73.7%，respectively(P<0.05).(2)The soil health evaluation results based on different datasets showed significant differences. By using the PLS-DA method，10 soil indicators were selected to construct the minimum dataset(including total carbon，organic carbon，total nitrogen，particulate organic carbon，particulate organic nitrogen，moisture content，pH，total phosphorus，carbon acquisition enzyme activity，and phosphorus acquisition enzyme activity)，combined with the nonlinear scoring method，the SHIPLSDA-NL coefficient calculated was the largest(17.7%)，and it had the highest correlation with the total dataset(R2=0.901，P<0.05). This indicated that the minimum dataset constructed based on PLS-DA was more suitable for replacing the total dataset to evaluate the soil health under different farming methods in the northeast black soil region，compared to the minimum datasets constructed based on principal component analysis and factor analysis.(3)According to SHIPLSDA-NL，the SHI value of no tillage with straw returning(0.50)was significantly higher than that of moldboard plowing with straw returning(0.45)and straw not returning(0.36)(P<0.05)，and showed the smallest difference，compared to the grassland reference(0.58). The minimum dataset constructed based on PLS-DA combined with the nonlinear scoring method could effectively evaluate the influence of different tillage methods on soil health in the northeast black soil region. No-till straw returning to the field significantly improved the health of farmland soil by enhancing carbon and nitrogen sequestration and increasing microbial activity.