Abstract: AIM: The study was undertaken to estimate acute toxicity in vivo using cytotoxicity data in vitro, in order to reduce animal usage in acute toxicity testing. METHODS: The basal cytotoxicity of seven compounds in Chinese hamster lung cell(CHL)cells were analyzed by MTT test, the LD50 values of acute toxicity were estimated by Registry of Cytotoxicity prediction model and validated by up-down method in mice. RESULTS: The IC50 values of compounds 1-7 in CHL cells were 0.43, 0.49, 0.18, 0.67, 3.03, 1.68 and 1.79 mg/mL, respectively. The predictive values of LD50 for compounds of 1-7 were 2376. 4, 2478. 3, 1574. 8, 2087.6, 4897.3, 3331.8 and 3300.7 mg/kg, respectively. The LD50 values for all compounds detected by up-down method were more than 2000.0 mg/kg except compound 4 with 1634. 0 mg/kg in female mice. Compared with the chemicals'toxicity classification based on predictive values and true values of LD50, the results indicated that toxicity classification of all compounds were basically at equal pace except compounds 3 and 4. CONCLUSION: The cytotoxicity data in vitro may be helpful in predicting acute toxicity in vivo and reducing the usage of laboratory animals.

Key words: new chemical entities, acute toxicity, cytotoxicity, RC prediction model