Spatial and temporal variations in physico-chemical properties of coastal waters play a major role in determining the density, diversity and occurrence of phytoplankton. The present study is conducted to assess spatio-temporal pattern of phytoplankton assemblage in the southern parts of the Caspian Sea (CS) which in turn can serve as an ecological health indicator in this region. Among 64 identified phytoplankton species, diatoms (37 species and 66.2%) and Dinophyceae (11 species and 18.2%) were dominant. Non-metric multidimensional scaling (NMDS) were employed to reveal Spatial and temporal distributions of phytoplankton assemblages. Four groups were established using hierarchical clustering based on species richness similarities which were absolutely represented the four seasons. The summer and spring samples were highly separated from the winter and autumn samples. The Linear Discriminant Analysis (LDA) model showed that temporal patterns of phytoplankton assemblages were mostly explained by chemical factors (silicate, phosphate, and nitrite) and temperature. In conclusion, results of this study suggested that spatio-temporal patterns of phytoplankton in the southern part of the CS are closely associated with seasonal variations in river flow and temperature. The temporal patterns are apparently dominant in this area and water nutrients are mainly responsible for seasonal changes.