Portal hypertension (PH) is the initial and main consequence of liver cirrhosis. Hepatic venous pressure gradient (HVPG) measurement has been widely used to estimate portal pressure gradient (PPG) and detect portal hypertension. However, some clinical studies have found poor correlation between HVPG and PPG, which may lead to the misdiagnosis of portal hypertension. In this study, we provided a method to evaluate patients’ PPG based on clinically measured HVPG with computational fluid dynamics (CFD). Twenty-five patients who underwent HVPG measurement were recruit for analysis. Results show that HVPG significantly correlates with PPG (R = 0.7499, P < 0.0001), with an accuracy to distinguish clinically significant portal hypertension (CSPH) as high as 92 %. However, PH severity classification was underestimated for 36 % patients, especially for patients with hepatic venous collateral formation and presinusoidal portal vein occlusion. It is concluded that HVPG is a relatively reliable diagnostic method for PH when PPG cannot be directly measured. For patients who have clinical symptoms of PH but their HVPG are within a normal range, numerical evaluation of PPG with CFD is an excellent way for their diagnosis.