Conventionally, a semi-quantitative microscopic nitroblue tetrazolium (NBT) assay is used to determine the production of superoxide anion (O2(-)) in various phagocytic cells. This microscopic assay is conducted by counting the cells containing blue NBT formazan deposits, which are formed by reduction of the membrane permeable, water-soluble, yellow-colored, nitroblue tetrazolium (Y-NBT) by O2(-). However, this assay is semi-quantitative and is prone to observer bias. In the present study, we modified the NBT assay by dissolving the blue formazan particles using 2M potassium hydroxide and dimethylsulfoxide and then measured its absorbance using a microplate reader at 620nm. The absorbance of dissolved NBT increased in proportion to cell number (r = 0.9907), incubation time, and stimulus concentration. To test the usefulness of this modified assay, we compared the abilities of a number of types of phagocytic cells to produce O2(-). The cells examined included murine macrophage cell lines (RAW 264.7 and J774), freshly prepared murine peritoneal macrophages and neutrophils, a human myeloid cell line (PLB-985), and freshly prepared human peripheral blood neutrophils. In addition, we demonstrate that nitric oxide produced by RAW 264.7 cells does not interfere with the modified colorimetric NBT assay. Taken together, our results indicate that the modified colorimetric NBT assay is simple, sensitive, and quantitative, and that it can be used to determine the amounts of intracellular O2(-) produced by phagocytic cells. Thus, this assay is sensitive enough to measure, quantitatively, even the small amounts of O2(-) produced in monocytes and macrophages that are not detectable by the conventional microscopic NBT assay.