Abstract:Kiwifruit ( Actinidiaspp.), belonged to Actinidiaceae, is a type of perennial deciduous woody liana and an important class of berry fruit. With rich sugar, protein, amino acids, vitamins and especially high vitamin C content, the kiwifruit is known as "the king of the fruit" and has a good market prospect. A. chinensisis a species endemic in China with a fast-expanding planting area due to its unique subtle flavor and high economic value. Optimization of planting scale and distribution of the crop has been the major concern for regional planning. The objective of this study was to test and determine the possibility of using the MaxEnt (the maximum entropy) model to simulate and predict future large-scale distribution of A. chinensis. Based on current environmental factors, three future climate scenarios suggested in the IPCC fifth report and current distribution sites of A. chinensis, we used the MaxEnt model in combination with ArcGIS to predict the potential geographic distribution and trend of change of A. chinensisin China. The dominant factors were chosen by using the Jackknife test and the Receiver Operating Characteristic (ROC) curve was used to evaluate the simulation. The results showed that high value of area under curve (AUC) denoted good results which significantly differed from random predictions. Based on the evaluation criterion, the accuracies of the predictions of A. chinensispotential distribution in the current and future periods were excellent. The predicted result of the MaxEnt model was imported into ArcGIS10.0 for further analysis and showed that under present climatic conditions, the total suitable area was 26.92% of the total land area in China. The potential distribution was highly consistent with the locations of specimen records and field surveys. The highly suitable areas were in Sichuan, Shaanxi, Chongqing, Hubei, Guizhou, Zhejiang, Hunan, Anhui, Henan, Jiangsu and Gansu Provinces. The areas of highly suitable habitat in the main producing provinces were analyzed statistically. The results showed that under the current conditions, the most suitable area for A. chinensiscultivation was 1.01×10 ⁶km ², accounting for 38.94% of the total suitable areas. The moderately suitable areas were in Henan, Hubei, Anhui and Shandong Provinces, with the area of 6.79×10 ⁵km ², accounting for 26.26% of the total suitable areas. Comparison of future suitable areas with current suitable areas showed that areas of high suitability increased under scenarios RCP2.6 and RCP4.5, but decreased under scenario RCP8.5. Under scenarios RCP4.5 and RCP8.5, the mean center of highly suitable area of A. chinensismoved northward. The result showed that the MaxEnt model was highly reliable in determining not only the range of geographic distribution of A. chinensis, but also in identifying dominant environmental factors driving the geographic distribution. Whereas climate was a decisive factor in species distribution, change in distribution pattern of species was the most direct effect of climate change. The results provided a critical reference base for A. chinensisplantation pattern and countermeasures to cope with climate change in China.