Abstract: Typhoons can generally bring the terrible damage for the agricultural production, particularly in Hainan Island with the frequent occurrence. It is a high demand for the accurate and rapid risk assessment on agricultural typhoon disaster, in order to optimize the layout of agricultural planting for the damage prevention and mitigation. Taking the Hainan Island of south China as the study area, this research aims to construct the assessment index system of agricultural typhoon disaster for the sustainable development countermeasures. Some influencing factors were also selected in the natural disaster system, including the rainfall, wind speed, typhoon occurrence time, digital elevation model data, rivers and watersheds, ratio vegetation index, agricultural population, agricultural output value, cultivated land area, and land area during the occurrence of typhoons over many years. Some considerations were then posed on the risk of disaster-causing factors, the exposure of hazard-formative environments, and the vulnerability of hazard-affected bodies. Eight indicators were illustrated for the risk comprehensive assessment model of agricultural typhoon disaster, including the maximum wind speed accumulation, rainfall accumulation, influence index of terrain factor, river network density, vegetation exposure, agricultural population density, output value per unit area of agricultural land, and the proportion of cultivated land in the land area. The relative importance of each index was scored by 15 experienced experts that engaged in the meteorology, agriculture, hydrology, and geography using analytic hierarchy process. Then, the weight was calculated to evaluate the contribution degree of eight indexes for each Criteria layer and index layer, according to the matrix judgment and consistency. As such, the agricultural typhoon disaster risk was determined from the typhoon occurrence data from 2001 to 2020. The agricultural typhoon disaster risks of cities and counties were divided into five levels: level I, II, III, IV, and V. The results showed that: The high risk of agricultural typhoon disaster was found in the study area, where the risk index was between 0.40 and 0.80, with an average of 0.61. The area of agricultural typhoon disaster risk level III and above were accounted for 74.8%. Spatially, the risk of typhoon disaster in the northern and coastal areas was much higher than that in the central mountainous area. Among them, Qionghai and Haikou cities presented the highest risk, with the disaster risk integrated index of 0.712 and 0.689, respectively. The lowest disaster risk was observed in Baoting, Baisha, Wuzhishan, and Qiongzhong cities, with the integrated index of disaster risk all lower than 0.540. The risk of disaster-causing factors and the exposure of hazard-formative environment for the agricultural typhoon were retained at a high level in northern areas (such as Haikou, Wenchang, Ding'an, and Chengmai), and coastal areas (such as Qionghai, and Danzhou), whereas a low level was found in the central areas (such as Baisha, Qiongzhong, and Wuzhishan). The vulnerability of hazard-affected bodies remained at a high level in Qionghai, Lingshui and Sanya, whereas, a low level in Baisha and Qiongzhong. Three countermeasures were also proposed for the prevention and reduction of agricultural typhoon disaster during sustainable development. Specifically, the layout of agricultural planting can be rationally adjusted for the risk monitoring and early warning, according to the typhoon characteristics and the risk distribution of agricultural typhoon disasters. The investment can be expected to strengthen in agricultural policy insurance. The specific measures included: 1) priority should be given to the typhoon-resistant crops, winter melon, and vegetable planting in the Qiongdong region; 2) an intelligent monitoring and perception network can be established using space-sky-terrestrial integration for the agricultural typhoon disaster risk in remote mountainous areas, while the new remote sensing can be introduced in the areas with developed agricultural activities; 3) to innovate the agricultural insurance underwriting and claim settlement mode. The finding can offer the strong support to the disaster prevention and emergency decision-making for the regions with different disaster risk levels.