Abstract: In the identification of the specimens collected from Zhoushan (Chushan) Islands of Zhejiang (Chekiang) Province, the authors have found that the species Aster tripolium L. (Tripolium vulgate Nees) of coastal salt marshes has a rather high amplitude of morphological variation. We, then, made a field investigation in Zhoushan Islands in October of 1980. A mass collection of five different populations of the same species of different ecological sites was carried out. Quantitative measurements of totally eight morphological characters, namely, the height of individual plants, the length and breadth of leaves and ligulate flowers, the number of lower branches, the degree of promineney of tap-roots, and the color of the base of stems, were made. The resulting data of these measurements and calculations were then compared by using methods of E. Anderson's pietorialized scatter diagram and standard deviation in order to bring to light the differentiation pattern of that species occurred in response to different habitats. Samples of soil of different population sites were also collected; their pH values and contents of NaCl were examined. According to the variation pattern of phenotypes as shown in these diagrams and figures, the populations of Aster tripolium of Zhoushan Islands could be roughly divided into the following two types: (1) the tall and long-leaved type of lower marshes with a soil salinity of 0.54%–0.56% of NaCl, including populations Nos. 1 and 2, with plants 49–51 cm. high (mean value), very few or no lower branches, usually dark purplish at the base of stems, and often without prominent tap-roots; (2) the low and short-leaved type of higher marshes and saline seepage areas above the limits of tidal submergence, with a soil salinity of 0.06%–0.17% of NaC1, including populations Nos. 3 and 4, with plants 19–26 cm. high (mean value), a few to manylower branches, and usually greyish white or slightly purplish at the base of stems. It is worth of note that these two different types of phenotypic variation of populations are on the whole corresponding to the two habitat types, which differ from each other chiefly in soil salinity resulted from the different levels of the sites, particularly the distance between the population sites and the salterns, and in the density of plants as well, which could probably be related to the intensity of competition. It is also interesting to note that owing to the site of population No. 5 being subjected to trampling, grazing and moving, the plants are usually mechanically stripped of their terminal buds. This is probably why the members of that population often show a peculiar pseudo- prostrate habit with their lower branches sometimes spreading at right angles to the main stem. However, the pattern of the population differinces obtained is often predetermined by the pattern of sampling. The real pattern of population differentiation of Aster tripolium is more likely clinal instead of abrupt. Without comparative experimental cultivation, it is difficult to decide whether the phenotypic variations mentioned above are genetically fixed or simply environmental modifications. Nevertheless, A. J, Gray et al. (1979) has shown that the large differences exist between populations of Aster tripolium at the different levels on the British salt marshes appear in uniform cultivation and may have a large genetic component. So, it might be expected that both of the factors-genetical and environmental, are likely to play their parts in the population differentiation of the said species. According to the classification of population systems given by V. Grant (1963), the two different types of populations of Aster tripolium of Zhoushan Islands may be considered as two ecological races.