The oxidation of 2',7'-dichlorodihydrofluorescein (2',7'-dichlorofluorescin, DCFH) to a fluorescent product, 2',7'-dichlorofluorescein (DCF), is commonly used to quantitatively measure oxidative stress in cells using a fluorescence microplate reader. However, many cell lines tend to grow non-uniformly in the wells. This non-uniform distribution results in a high degree of variability in the fluorescence signal and decreases the precision of the method. Also, samples treated in large culture plates, dishes or flasks cannot be assayed directly in fluorescence microplate readers. This study reports an improved DCF assay method that lyses cells with DMSO/PBS (90% dimethyl sulphoxide/10% phosphate buffered saline). Oxidative stress was induced with either hydrogen peroxide or an hypoxia-reoxygenation treatment. Cell lysis with DMSO/PBS resulted in highly stable fluorescence signals in comparison to Triton X-100/PBS lysed cells. The precision of DCF fluorescence measurements of DMSO/PBS lysed cells was much better than for attached cells measured directly in 96-well plates. While DCF fluorescence in PBS was strongly quenched by albumin, no quenching occurred in DMSO/PBS. In conclusion this study describes a more convenient and accurate method for measuring cellular oxidative stress that also makes it possible to assay cells treated in large culture plates.