The aim for the present study was to analyze data from four otter trawl surveys (conducted between 2015 and 2016) in order to investigate variation in the composition and diversity of fish species in Daya Bay, South China. The data included 131 species, which belonged to 84 genera, 53 families, and 14 orders. Species of the Perciformes were the most predominant, accounting for 54.20%, followed by those of the Agunilliformes and Pleuronectiformes, each accounting for 9.92%. Seasonal (temporal) variations of species number were obvious, with the greatest number of species observed during summer (=32). The predominant fish species included Clupanodon punctatus, Leiognathus brevirostris and . Diversity analysis revealed significant seasonal and spatial differences in fish species diversity. The diversity index (') ranged from 1.516 to 1.998, with the greatest value observed during winter and the lowest during autumn, whereas the evenness index (') ranged from 0.494 to 0.869, with the greatest value observed during winter and the lowest during summer, and the richness index (') ranged from 2.230 to 3.777, with the greatest observed during summer and the lowest during autumn. Model analysis indicated that temperature, salinity, and water depth are the main environmental factors affecting the structure of the fish community in Daya Bay. In terms of spatial distribution, fish diversity was greater in the middle of the bay than in coastal waters. Compared with historical data, the fish community structure has changed significantly, as a result of disturbance from anthropogenic activities, including obvious changes in the composition of dominant species and reduced diversity, suggesting the simplification of fish species appeared in this bay.