To develop benthic macroinvertebrate-based multimetrics for the bioassessment of the water quality of Taihu Lake, benthic macroinvertebrate assemblages were sampled at 60 sites in 8 subareas of Taihu Lake. A total of 24 species of the benthos were collected: 14 species of Mollusca, 5 species of Arthropoda and 5 species of Annelida. Based on Hilsenhoff biotic index discriminatory criteria, 15 sites were considered relatively clean while the remaining 45 sites were polluted . A comprehensive evaluation of the metrics was carried out. The results showed that 7 metrics (i.e., total number of taxa, the number of Mollusca & Crustacea taxa, % Mollusca & Crustacea, % Gastropoda, Goodnight-Whitley index, Hilsenhoff biotic index and % Collector-gathers) were sensitive. A score of 5, 3 or 1 was assigned to every metric for standardization of the metrics, and a multimetric on a scale ranging from 7 to 35 for bioassessment was developed for each site by addition of its 7 metrics. Four levels of discriminatory biocriteria for water quality were eventually proposed by quartation: 7-14, very poor; 15-21, poor; 22-28, fair; and 29-35, good. The re-scored results from the 60 sites showed that the water quality of southeastern Taihu Lake was relatively good, whereas that of northwestern Taihu Lake was relatively poor. Basically, the discriminatory biocriteria are suitable for the assessment of the water quality of Taihu Lake.