Aim This study tested a new criteria for electrocardiogram to improve the accuracy of primary hypertension with left ventricular hypertrophy. Methods The study measured left ventricular mass according to the American Society of Echocardiography (ASE) standard. We selected 129 patients with hypertension (HBP) and left ventricular hypertrophy (LVH) diagnosed in our hospital. At the same time, the normal wall thickness group of hypertension was matched according to age and gender. It simultaneously recorded 12-lead ECG, selected the deepest S-wave (SD) of all leads as the research object, and currently accepted LVH ECG criteria such as Cornell voltage and Sokolow-Lyon were used for comparison. Area under the curve (AUC) analysis was used for comparison of single and combined leads. Results The AUC, sensitivity, and specificity of SD predicted HBP with LVH were 0.2,6.05% and 81.65%, and SD+Sv4 showed the highest sensitivity of 88.37%. The equivalence test showed that Z value of SD was statistically different from that of Cornell,Sokolow-Lyon and SD+Sv4(all P＜0.05). The AUC, sensitivity and specificity of predicted male HBP with LVH were 0.1,0.29% and 75.34%. The equivalence test was superior to Cornell, Sokolow-Lyon and SD+Sv4 (all P＜0.05). Conclusion The new criteria SD to diagnose HBP with LVH improves the sensitivity of diagnosis, superior to Cornell and Sokolow-Lyon standards which is worthy of clinical promotion.