The intersystem crossing and dispersive electrontransfer dynamics of eosin Y (EY) photosensitizers are probed using single-molecule microscopy. The blinking dynamics of EY on glass are quantified by constructing cumulative distribution functions of emissive ("on") and nonemissive ("off") events. Maximum likelihood estimation (MLE) and goodness-of-fit tests based on the Kolmogorov-Smirnov (KS) statistic are used to establish the best fit to the blinking data and differentiate among competitive photophysical processes. The on-time probability distributions for EY in N-2 and air are power-law distributed after similar to 1s, with fit parameters that are significantly modified upon exposure to oxygen. By extending the statistically principled MLE/KS approach to include an onset time for log-normal behavior, we demonstrate that the off-time distribution for EY in N-2 is best fit to a combination of exponential and log-normal functions. The corresponding distribution for EY in air is best fit to a log-normal function alone. Furthermore, power law and log-normal distributions are observed for an individual molecule in air, consistent with dynamic fluctuations in the rate constant for dark-state population and depopulation. These observations support the interpretation that dispersive electron transfer (i.e., the Albery model) from the first excited singlet state (S-1) of EY to trap states on glass is predominately responsible for blinking in oxic conditions. In anoxic environment, both triplet-state blinking and dispersive electron transfer from S-1 and the excited triplet state (T-1) contribute to the excited-state dynamics of EY.