From a morphological point of view, we attempted to estimate rice (Oryza sativa L.) yields by multiplication of four yield components considering seasonal changes in meteorological conditions. We focused on the rice cultivar Kirara397, for which large amounts of cultivation test data were available. The multiple linear regression method was applied for modeling between each yield component of Kirara397 and meteorological data. The compensation effect between yield components of rice was also considered. For other varieties (Nanatsuboshi as a panicle weight-type cultivar and Fukkurinko as a panicle number-type cultivar, relative to Kirara397) with less cultivation test data available, we formed the following two assumptions: 1) the type of weather elements affecting the yield components would be the same, regardless of the rice variety; and 2) the ratios of each yield component among rice varieties would be constant under the same weather conditions. Therefore, yield components for these varieties were estimated using correction ratios determined previously with yield component data. The determination coefficients for brown rice yield of Kirara397, Nanatsuboshi, and Fukkurinko were 0.440, 0.698, and 0.669, respectively. This method was thought to be a convenient and useful method for evaluation of yield and four yield components with only two meteorological data (daily air temperature and solar radiation) and dates of rice development. However, both our assumptions for rice cultivars with a small number of cultivation data were not fully adequate for estimating the yield and four yield components quantitatively, while it could be useful for determining relative quantities, particularly for panicle number, thousand kernel weight, and brown rice yield.