Photobleaching and related photochemical processes are recognized experimental barriers to quantification of fluorescence by microscopy. We have measured the kinetics of photobleaching of fluorophores in living and fixed cells and in microemulsions, and have demonstrated the spatial variability of these processes within individual cells. An inverted fluorescence microscope and a high-sensitivity camera, together with high-speed data acquisition by a computer-controlled image processor, have been used to control precisely exposure time to excitation light and to record images. To improve the signal-to-noise ratio, 32 digital images were integrated. After correction for spatial variations in camera sensitivity and background fluorescence, the images of the relative fluorescence intensities for 0.065 micron2 areas in the object plane were obtained. To evaluate photobleaching objectively, an algorithm was developed to fit a three-parameter exponential equation to 20 images recorded from the same microscope field as a function of illumination time. The results of this analysis demonstrated that the photobleaching process followed first-order reaction kinetics with rate constants that were spatially heterogeneous and varied, within the same cell, between 2- and 65-fold, depending on the fluorophore. The photobleaching rate constants increased proportionally with increasing excitation intensity and, for benzo(a)pyrene, were independent of probe concentration over three orders of magnitude (1.25 microM to 1.25 mM). The propensity to photobleach was different with each fluorophore. Under the cellular conditions used in these studies, the average rates of photobleaching decreased in this order: N-(7-nitrobenz-2-oxa-1,3-diazole)-23,24-dinor-5-cholen-22-amine-3 beta-ol greater than acridine orange greater than rhodamine-123 greater than benzo(a)pyrene greater than fluorescein greater than tetramethylrhodamine greater than 1,1'dioctadecyl-3,3,3',3'-tetramethylindocarbocyanine. The photobleaching appears to be an oxidation reaction, in that the addition of saturated solutions of Na2S2O5 to mineral oil microemulsions eliminated photobleaching of N-(7-nitrobenz-2-oxa-1,3-diazole)-23,24-dinor-5-cholen-22-amine-3 beta-ol or benzo(a)pyrene. We identified experimental conditions to observe, without detectable photobleaching, fluorophores in living cells, which can not be studied anaerobically. Useful images were obtained when excitation light was reduced to eliminate photobleaching, as determined from zero-time images calculated from the exponential fit routine.(ABSTRACT TRUNCATED AT 400 WORDS)