High blood pressure is a major risk factor for cardiovascular disease， and frequent use of blood pressure medications can have side effects. Therefore， safe and efficient food-borne antihypertensive peptides have become the focus of attention. In this study， a quantitative structure-activity relationship(QSAR) model of angiotensin-converting enzyme (ACE) inhibitory peptide was established to determine the relationship between the structure and activity of ACE inhibitory peptide. The model results were combined with computer virtual enzyme digestion to determine proteolytic sites accurately. Suitable proteases were selected to hydrolyze hemp seed protein to prepare ACE inhibitory peptide， and the inhibitory effect of the final hydrolysate was determined. The results showed that the SVM-AAindex model based on dipeptide had the best predictive performance(R2 0.81， RMSE 0.53)， and had a strong ACE inhibition effect when the C-terminal amino acids of dipeptide were hydrophobic and aromatic amino acids. The analysis of protein components showed that hemp seed protein contains 34.60% hydrophobic amino acids and aromatic amino acids， which is a good source for preparing ACE inhibitory peptides. The alkaline and subtilisin， which produced more short peptides and bioactive peptides by virtual enzyme digestion， were selected for directed hydrolysis of hemp seed protein. The ACE inhibition values of the final hydrolysates were 1.89 and 2.30 mg/mL， respectively. In this study， an efficient and accurate method for preparing ACE inhibitory peptide from hemp seed protein was established， which has a good application prospect in antihypertensive drugs.