A process-based model was used to simulate a hindcast based on the worst historical water quality condition of a tropical urban reservoir. Paranoá Lake is located in Brasília-DF, Brazil, and went through intense eutrophication in the 70s and 90s, with an important cyanobacterial bloom event in 1978. The parameters of phytoplankton were calibrated, focusing on the group of Chlorophyta (green algae) and Cyanobacteria (blue-green algae) at four depths (1, 10, 15 and 20m). The results indicated that the model was able to reproduce the Cyanobacteria biomass in comparison with the observations (RMSE=22-29.10-3 mgC L-1). On the other hand, the simulated Chlorophyta biomass showed good agreement with the observed data only in the bottom layer (RMSE=29. 10-3 mgC L-1 at 20m). In the hindcast simulation, the model was able to predict a significant increase in cyanobacterial biomass facing a water quality deterioration. In the meantime, the simulated Chlorophyta biomass decreased, which may indicate the phytoplankton group succession in response to the environmental conditions.