In this study, I use statistical models to find spatial patterns in long-term temporal trends of vegetable yield in Japan. I analyze changes in annual yields of cabbage, napa cabbage, and lettuce over the period of 1973–2016 years in the 47 prefectures of Japan by using the data of agricultural dataset. The yield trends for each prefectures are classified into four categories including increasing, stagnating, collapsed, and never improved, by using regression models. I then map these different temporal trends in prefectures and discuss their spatial patterns for the vegetables. The annual yields were increasing or stagnating in more than half of the prefectures in Japan for cabbage and lettuce. The annual yield was collapsed in more than half of the prefectures for napa cabbage yield. For cabbage and lettuce, yield-collapsed prefectures were mostly along the coast of the Sea of Japan. On the other hand, for napa cabbage, yield-collapsed prefectures were around the country. By analyzing spatial patterns in long-term temporal trends of vegetable yield, this work may provide strategies and investments to increase or maintain agricultural production in Japan which meet food demands.