Abstract: The continuous flow technique coupled with GasBenchⅡ-Isotope Ratio Mass Spectrometry (IRMS) became the routine method to analyse the stable carbon and oxygen isotope compositions of carbonate. Previous studies revealed that the isotope results were influenced by the varied experimental conditions. However, rare studies have assessed the influences of multiple experimental conditions on the isotope results. Here, all reaction conditions including flushing time, reaction temperature, reaction time and chromatographic separation temperature have been evaluated systematically. Flushing time longer than 9 minutes can eliminate the interference of air. Different reaction temperatures and time have distinct influences on the isotope values. 72℃, 60 minutes and 60℃ have been chosen for the reaction temperature, reaction time and chromatographic separation temperature respectively. The analytical precisions are better than 0.03‰ (for δ¹³C) and 0.05 (for δ¹⁸O) under the optimized conditions. In addition, different isotope normalization methods would affect the isotope results. Two isotope normalization methods have been compared, based on about 4000 samples of isotope data. The calculated results from two standard samples are much more precise and reliable than from single standard samples. The study provides a reference for experimental conditions in GasBench Ⅱ-IRMS technology, ensuring the accuracy and reliability for carbon and oxygen isotope analysis. At the same time, the experimental conditions should be further investigated in real samples analysis due to the inhomogeneous and complex sample compositions.