Abstract: Carbon-nitrogen ratio (C/N) and carbon-sulphur ratio (C/S) of unknown samples can vary greatly in the natural environment. There are also distinct differences in the ionization efficiencies of N and S elements. These factors will introduce the fluctuation of signal intensity to the element analysis-isotope ratio mass spectrometry (EA-IRMS), which can has a further impact on the measurements of sample δ¹⁵N and δ³⁴S. However, the analysis of unknown sample δ¹⁵N and δ³⁴S is always restricted by the low content or the scarcity of the target elements. In this study, we focused on the effect of sample amount and signal intensity on the δ¹⁵N and δ³⁴S measurement by EA-IRMS. Our results indicated that the EA-IRMS measurement of δ¹⁵N and δ³⁴S was significantly affected by the “blank effect” when the sample amount and signal intensity were limited during the mass spectrometry analysis. Under the injection conditions of 1.71-561.93 μg N and 3.74-100.62 μg S, the variation ranges of the measurement errors of the isotopic reference materials δ¹⁵N and δ³⁴S were 0.18‰-0.54‰ and 0.21‰-0.88‰, respectively. Meanwhile, the variation ranges of the measurement errors of measurement precision were 0.24‰-0.57‰ and 0.27‰-0.54‰, respectively. With the high C/N and C/S wood samples, the measurement errors of EA-IRMS were 0.34% for δ¹⁵N. Because the true value of the above wood sample δ³⁴S was unknown, the measurement precision of EA-IRMS on the wood samples was 0.58‰. Based on the analysis of two wood sample δ15N, the EA-IRMS had a good time stability during the continuous (6-20 days) measurement process. We hoped these findings could provide implications for the application of EA-IRMS in biogeochemistry and life science areas.