In this study, the biological and chemical diversity of 8 symbiotic marine fungal strains, with strong cytotoxicity against brine shrimp larvae, were investigated by nucleotide sequencing, morphology and cluster analysis of HPTLC fingerprint. These strains were identified by ITS rDNA sequencing, phylogenic analysis, and morphology to be Hypocrea lixii, Chaetomium globosum, Aspergillus fumigatus, Asp. clavatus and Alternaria sp. Their differences in secondary metabolites were shown by cluster analysis of digitalized colors of HPTLC spots, a newly developed method, which produced a similar dendrogram with that of ITS cluster analysis. Furthermore, this method can fully display intraspecific differences and even the remarkable difference in Aspergillus strains which goes beyond the boundary between genera. Their biological-chemical diversity may be the basis of their potent cytotoxicity and implies their potential in producing diversified antitumor or pesticidal constituents. Symbiotic fungi live on the surface or in the inner tissue of their hosts. Some terrestrial symbiotic fungi have been found to produce toxins or anti-feedants to protect their hosts from predators and grazers [1]. Some of these compounds can be used as antitumor or pesticidal agents. Symbiotic marine fungi have been isolated from seaweeds, sponges, corals, mangroves and sea grasses, also showing taxonomicall diversity and producing numerous active compounds [2]. The intertidal coastline of Dalian possesses diversified natural and artificial habitats and also high biodiversity of marine plants, invertebrates, and microorganisms. In our screening for useful cytotoxins from local symbiotic marine fungi using brine shrimp lethality test, a widely used bifunctional preliminary screening model to discover antitumor drugs and pesticides from the sea [3][4][5], eight strains with potent activities were discovered. Herein, we report the study on the biological and chemical diversity of these bioactive strains by ITS rDNA sequence analysis, morphology, and metabolite fingerprinting using a new cluster method.