:Saccharomyces cerevisiae is a model organism for aging and longevity studies. In a clonal population of S. cerevisiae, the timing of cell death in the stationary phase is not synchronized, indicating that heterogeneity exists in survival at a single-cell level. Heterogeneity also exists in the cell size, and its correlation with the death rate has been discussed in past studies. However, the direct cause of the heterogeneity in survival remains unknown. In this report, we revisited this question and asked whether the death rate has any correlation with cell size. Past studies did not exclude a possibility that cells change their size upon or after death. If such a change exists, the size dependence of cell death could be misinterpreted. Therefore, we analyzed the correlation between the death rate and cell size before death by time-lapse imaging. It turned out that the size dependence of the death rate varied from one strain to another, suggesting that general principles between cell size and death do not exist. Instead, cells shrink upon cell death, resulting in the accumulation of small dead cells. The degree of cell shrinkage was proportional to the cell size, and the ratio was constant in two strains, which is between 25 and 28%, suggesting the presence of general principles and mechanisms behind the shrinkage event upon cell death. Further investigation of the cause and mechanism of the shrinkage will help us to understand the process of cell death and the origin of the heterogeneity in survival. IMPORTANCE Cells display various behaviors even though they originate from a clonal population. Such diversity is also observed in cell survival in the stationary phase of Saccharomyces cerevisiae. However, we know little about the causes of heterogeneity in the timing of cell death at a single-cell level. To deepen our understanding of the cause of heterogeneity, we observed the process of cell death in S. cerevisiae by time-lapse imaging. Our analysis showed that cells shrank upon cell death, resulting in the accumulation of small dead cells, while a general principle in the correlation between cell size and death was not seen. The degree of cell shrinkage was proportional to cell size before cell death, and it was constant under all conditions tested, indicating the presence of general principles behind the shrinkage event. Future studies to identify the cause of cell shrinkage must contribute to finding the origin of the heterogeneity in survival.