A long-term dataset, including physicochemical, nutrient, and phytoplankton assemblages from 1994 to 2016, was analyzed to investigate the response of the algal community to variations in environmental factors in Deep Bay and Mirs Bay in southern China. These bays differ in their overall nutrient loadings, as well as in physical factors. The results showed that diatoms were numerically dominant in Mirs Bay, while other minor phytoplankton groups, including eutrophication-tolerant species, constituted the majority in Deep Bay. Phytoplankton community composition tended to be less complex in Deep Bay, suggesting a stressed, unstable and unbalanced ecosystem compared to that in Mirs Bay. Algal blooms occurred more frequently in Mirs Bay, whereas fewer but larger-scale blooms occurred in Deep Bay. Statistically, the combination of all explanatory variables accounted for approximately 55% of the variation in Chlorophyll-a (Chl-a) concentration and less than 20% of the total phytoplankton variation over the 23-year period in the two bays. The high level of nutrients caused by urbanization was not the driving force in the formation of blooms but presumably provided a nutrient base that resulted in blooms with longer durations and covering larger areas.