Risk analysis has been used regularly in the management of aquatic animal health, resulting in increased success at preventing and controlling diseases. A number of factors affect the efficacy of a vaccine in aquaculture. Thus, it is important to identify these risk factors and estimate their weights. We developed a quantitative risk assessment model for the prevention for grass carp hemorrhage disease by analyzing epidemiological study data using the Delphi method and meta-analysis. The model was composed of a risk hierarchy system, weight and valuation standards for each risk factor, and an integrated evaluation function. The risk hierarchy system included 5 criterion layers and 24 index layers. The criterion layers included vaccine quality (B1), immune program (B2), fish (B3), pond environment (B4), and care and management (B5) with a set of weight about ={0.267, 0.102, 0.131, 0.263, 0.237}. The index layer consisted of vaccine species(C1), storage temperature and expiration date(C2), storage and transportation conditions (C3), storage of the vaccine (C4), health status of the fry following vaccination(C5), vaccination technique(C6), vaccination dosage(C7), number of unimmunized fry (C8), source of fry (C9), health status of fish(C10), size of fish(C11), water temperature(C12), dissolved oxygen(C13), ammonia nitrogen(C14), nitrite(C15), pH (C16), water transparency(C17), water color(C18), thickness of mud(C19), rearing density(C20), culture type(C21), drug use(C22), feed(C23), and grass feed(C24) with a set of absolute weights about ={0.138, 0.059, 0.046, 0.024, 0.035, 0.018, 0.027, 0.022, 0.040, 0.054, 0.037, 0.076, 0.037, 0.032, 0.030, 0.027, 0.018, 0.019, 0.024, 0.104, 0.024, 0.027, 0.062, 0.020}. The results of the risk assessment were represented by the integrated evaluation function. Using vaccination data from southwest China to validate the model, the integrated evaluation value from three infected ponds was 0.572, 0.638, and 0.617. These values were significantly higher than values from non-infected ponds (<0.01).The results of the risk assessment are consistent with the survey data. Thus, our model can be used to reduce the risk of grass carp hemorrhagic disease by allowing culturists to plan with more certainty for increased immune efficacy.