Climate change may induce severe frost damage to crops, and thus a reasonable assessment of frost risk, considering both crop phenology and meteorology, is required. Here, we introduced a new index of potential frost risk F risk using thermal time minimum air temperature below the threshold value weighted by the percentage of budburst P bud . Moreover, we evaluated the spatiotemporal distributions of F risk in tea fields within a 60 km 60 km area in east Japan from 1981-2020, using 1 km 2 -gridded meteorological data and a newly developed model of P bud . The P bud model considered three phenological phases endodormancy, ecodormancy, and progress of budburst and successfully represented changes in the P bud of the tea buds for 15 years, with root mean square errors of 8.5 percentage points. The spatiotemporal distributions of F risk over the past 40 years showed that potential frost risk significantly increased at elevations ranging from 50 m to 300 m because the budburst advanced at a faster rate than the temperature warming. These elevations corresponded to areas where tea plants were mainly cultivated, which indicates that tea cultivation is becoming vulnerable to frost, and the risk of economic losses due to the frost is increasing. The proposed assessment of frost risk could contribute to predicting frost damage and developing more reliable strategies for the operation of frost protection under the effects of future climate change.