Abstract: OBJECTIVE To establish and verify a detection method for bacterial endotoxins using the micro-kinetic turbidimetric method. METHODS The dynamic turbidimetric limulus reagent was used, the sample volume of each well and the limulus reagent 25 μL was used, the detection wavelength was 405 nm, the onset OD value was 0.03, and the half-well microtiter plate was used. The accuracy, reproducibility, linearity, range, and durability of the method had been verified by 4 laboratories collaboratively. A total of 6 vaccines included recombinant novel coronavirus vaccine, recombinant novel coronavirus vaccine(type 5 adenovirus vector), recombinant hepatitis B vaccine, combined attenuated vaccine for measles and mumps, inactivated bivalent hemorrhagic fever with renal syndrome vaccine, and influenza vaccines were being studied for method applicability. RESULTS The total average recovery rates of 24 trials in 4 laboratories in the high, medium and low dose groups were 100.8%, 150.0% and 125.3%, respectively, the RSD values were 5.3%, 11.0%, and 20.7%. The standard curve had an average correlation coefficient of 0.990 in the range of 0.01-10 EU·mL^-1. The durability study was investigated. The influence of the two sources of limulus reagent on the detection was in compliance with the regulations. The recovery rates of the interference test for a total of 18 batches of samples of 6 varieties were all within between 50% and 200%, the interference test met the requirements. CONCLUSION A method for detecting bacterial endotoxins using micro-kinetic turbidimetry has been established. This method has good accuracy, reproducibility, linearity and range, durability, and has good variety applicability among 6 biological products. This method can be used for bacterial endotoxin detection of conventional samples.