Abstract: In order to understand the diversity of Chaetoceros lorenzianus complex and divide C. decipiens and C. lorenzianus clearly, Guangdong coast was chosen as a case studied area and a total of 22 monoclonal strains within C. lorenzianus complex were established in this study. Using optical and electron microscopy, their morphology was observed through life cycle, including vegetative cells and resting spores. Hypervariable D1-D3 region of nuclear ribosomal large subunit encode gene for each strain was amplified and their molecular systematics was studied. Following conclusions were made based on the combination of morphology and molecular data. The results showed that the fusing setae base was not a steady feature and could not be considered as an exclusive characteristic for C. decipiens, which supporting the opinion from Chen et al. (2017). The strains with similar morphology with C. lorenzianus located into two clades, and it indicated the presence of unexplored species in C. lorenzianus complex. The strains in one clade shared most similar morphology with present-known C. lorenzianus, and those in another clade were distinct in the morphologies of setae poroid and resting spore. When compared wtih C. lorenzianus descriptions in previous studies, confusion and disorder for C. lorenzianus delineation were found. For example, there were contradictory and competing descriptions on the shape of resting spore and the presence of dichotomous branches on the primary valve. In future studies, the re-examination on type C. lorenzianus materials, including the vegetative cells and resting spores, is proposed to be conducted preferentially, and then the true diversity might be achieved.