Diatoms are photosynthetic microorganisms of great ecological and biogeochemical importance, forming vast blooms in aquatic ecosystems. These blooms are characterized by proliferation of the population followed by a rapid demise due to diverse environmental stressors. However, it remains unclear how some diatom cells survive the demise and serve as a seed for the next bloom, while most do not. Current understanding of phytoplankton acclimation to stress is based on population-level analyses, masking cell-to-cell variability. Here we investigated heterogeneity within isogenic diatom populations in response to oxidative stress, which mediates a wide range of environmental stress conditions. We combined flow cytometry and a microfluidics system for live-imaging microscopy to measure redox dynamics at the single-cell level. Using the redox-sensitive sensor roGFP, we measured in vivo oxidation patterns in the model diatom Phaeodactylum tricornutum. Chloroplast targeted roGFP exhibited a light-dependent, bi-stable oxidation pattern in response to H2O2 and high light, revealing two distinct subpopulations. The oxidized subpopulation was sensitive to the stress and subsequently died, while the reduced subpopulation survived. Oxidation of the chloroplast glutathione pool preceded commitment to cell death, and was used as a novel predictor of cell fate. Transcriptome analysis of subpopulations sorted based on chloroplast roGFP oxidation revealed genes involved in cell death and survival, which are currently under an ongoing investigation. These findings suggest that light modulates the sensitivity of diatoms to environmental stressors. We propose that phenotypic variability within diatom populations can provide an ecological strategy to cope with rapid environmental fluctuations in the marine ecosystem.