Algal blooms have been increasing in frequency in recent years in Xiamen seas, primarily in the western Xiamen Harbour, with occurrence in the spring and summer almost every year. There have been extensive reports on the phytoplankton in Xiamen seas，but few deals with the characterization of population dynamics of the blooms. The present study analyzes the community structure of the phytoplankton collected on 10 cruises from 11 stations during the summer red tide of 2000 (June to July) in order to understand the dynamics of phytoplankton assemblages under pollution stress conditions. The results show that: (1)the spatial and temporal variations of the phytoplankton diversity index (H′) are large during the survey period，ranging from 0.45 to 4.23; (2) based on the results of the multivariate analysis (cluster analysis and multi-dimensional scaling), the phytoplankton assemblages can be divided into three groups associated with the specific stages of phytoplankton succession and with the different physic-chemical features associated with each sampling region. Group I was composed of the samples collected from the low salinity stations in western Xiamen Harbour in June, 2000. The community was consisted of nanophytoplankton flora including the Bacillariophyta (such as Chaetoceros socialis, Skeletonema costatum) and the Chrysophyta (such as Chromulina sp.). These species were responsible for the red tide. Groups Ⅱ and Ⅲ were both dominated by microphytoplankton assemblages. Group II was represented by the subtropical coastal eurytopic species commonly found in the studied area before and after red tide blooms, such as Paralia sulcata，Proboscia alata，Thalassionema nitzschioides and Asterionellopsis glacialis. Group Ⅲ included mainly the samples collected from the stations in the eastern-northeastern Xiamen seas，represented by neretic species that are ubiquitous in the coastal areas of southeast China, including Chaetoceros curvisetus，Leptocylindrus mediterraneus，Pseudo-nitzschia pungens，Schroederella delicatula and Guinardia striata; (3) in the Abundance-Biomass Comparison analysis, the K-dominant curve for numerical abundance lies below that for the biomass when the nanophytoplankton bloom occurred, while both curves largely coincided during non-bloom periods. Such a pattern may indicate the heavily polluted condition in the western waters and the moderately polluted conditions in the eastern-northeastern waters in the Xiamen seas; 4) based on the results of principal component analysis, the first three components accounted for 78.5% of the total variance and the most important factor loadings include species richness, dissolved oxygen, NH⁺₄-N and PO^-₄-P, salinity, and temperature.