Algal blooms constitute an emerging threat to global inland water quality. As one of the biggest lake and important water resource in the world, Lake Victoria is facing recurrent proliferation of water hyacinth and cyanobacteria. To better manage and improve water resources, the spatiotemporal distribution and long-term trends of algal blooms must be understood, as well as the driving factors. In this study, we used more than 20 years of Landsat series images to extract and map algal bloom occurrences based on a neural network model in the Transform framework and to revel the long-term changes law and trends of cyanobacterial. Results showed that both the bloom occurrence frequency (BOF) and the affected areas exhibited an increasing trend with the rates of 2.87%∙yr^−1and 981 km²∙yr^−1, respectively. Some crucial driving factors were selected to analyze climatic and anthropogenic impact on bloom occurrences. For meteorological factors, lake water volume has shown positive correlation with BOF (r equals 0.58), while precipitation is positively correlated with BOF variations (r equals 0.57), with 1-year lag. The annual precipitation range that contributes to BOF increase in Lake Victoria was estimated to be approximately 98-118 km³∙yr^−1. For anthropogenic factors, socio-economic development, expansion of built-up areas and croplands, and decrease in ecological land areas, such as wetland and grassland, largely contributed to the BOF increase in Lake Victoria. Based on the above results, the degree of influence of the factors was analyzed using grey relational analysis (GRA), with lake water volume and socio-economic being the predominant driving factors. This study provides valuable insights into the long-term algal bloom occurrence dynamics in Lake Victoria, and could provide important data support for the ecological safety and sustainable use of the lake.