Abstract:To explore the edible safety of Rhizobium pusense，acute and subchronic oral toxicity tests were performed for toxicological safety assessment in this study. A limit dose method was adopted，and acute oral toxicity detection was carried out on 20 SD rats and 20 KM mice at a gavage dose of 10.0g/kg. For the subchronic oral toxicity test，80 SD rats were divided into three dose groups （2.5，5.0，10.0mL/kg） and one control group，and all rats were treated with continuous oral gavage for 90 d. Another 20 SD rats were assigned to the satellite test，including a satellite control group and a satellite high-dose group. After 90 d of gavage administration，the test substance was withdrawn，and the rats were fed normally for another 28 d for continuous observation. The results showed that the acute oral median lethal dose （LD50） of Rhizobium pusense for SD rats and KM mice was higher than 10g/kg. No toxic symptoms or deaths occurred in SD rats during the subchronic test. No dosedependent statistical differences were observed in routine blood parameters，blood biochemistry，serum electrolytes and blood coagulation indices. Bladder pathological lesions，intestinal gland hyperplasia and elevated urinary protein were detected in the high-dose group，among which bladder lesions and urinary protein elevation mainly appeared in female rats. Only intestinal gland hyperplasia was found in the satellite high-dose group.No irreversible adverse pathological changes and abnormal organ coefficients were observed in all treatment groups. In conclusion，Rhizobium pusense was practically non-toxic under the experimental conditions. The no observed adverse effect level （NOAEL） of Rhizobium pusense in the 90-day subchronic oral toxicity test was≤10mL/kg.