Objective To evaluate and compare the application value of predictive models of unilateral primary aldosteronism (PA) in Chinese population. Methods Two hundred fourteen patients diagnosed with PA and completed adrenal venous sampling (AVS) were included. The patients were subtyped based on blood collection results. There were 129 patients with unilateral PA and 85 with bilateral PA. Based on the subtype, the diagnostic value of the 2016 Endocrine Association Guidelines for PA and 6 IHA prediction models (Kupers score model, Nanba score model, CONPASS model, SPACE score model, Puar model and Ma model) was evaluated using receiver operating characteristic (ROC) curve and sensitivity and specificity analysis. Results Only 3 patients met the 2016 American Primary Aldosteronism Guideline criteria and could be directly diagnosed as unilateral PA without AVS. Its specificity was 100% , and the sensitivity was only 2． 33% . The areas under the ROC curves (AUC) of the other models were less than 0． 7. The sensitivity was between 16． 28% and 71． 26% , and the specificity was between 37． 74% and 86． 30% according to the cut-off point recommended by the authors. The total score of SPACE model confirmed by captopril inhibition test was ≥19, the specificity of unilateral PA was 98． 63% and the sensitivity was 9． 17% . Conclusions For safety, only patients who met 2016 American guideline criteria for PA and those with a total SPACE score ≥19 confirmed by captopril challenge test would bypass AVS and proceed to surgery. Most patients still need AVS to guide the next treatment plan.