RATIONALE In recent experiments, we found that compound-specific d2H values can vary as a result of changing the gas chromatography temperature program under common pyrolysis conditions. To achieve better precision, it is necessary to examine the details and find a solution to this problem when using gas chromatography/thermal conversion/isotope ratio mass spectrometry (GC-TC-IRMS) for hydrogen isotope analysis. METHODS A test was designed to find the possible temperature effect under four different GC temperature ramp rates using n-alkanes (n-C21, n-C27, and n-C31) and fatty acids (n-C12, n-C18, and n-C24). The common 'hexane' method was used initially to condition the pyrolysis reactor. Experiments were then carried out using the 'methane condition' method because it was considered to improve pyrolysis efficiency. RESULTS Under the 'hexane condition' the measured hydrogen isotope ratios of the n-alkanes and n-fatty acids became more positive with increasing GC temperature ramp rate. The ion current intensity of hydrogen also generally increased. However, when the 'methane condition' method was used, the measured d2H values of the n-alkanes and n-fatty acids showed little change under different GC temperature ramp rates. CONCLUSIONS Higher pyrolysis efficiency could reduce the tailing of the H2 peak and the related isotopic variations at increased GC temperature ramp rates. In addition, too slow a temperature ramp rate could broaden the peak width and thus increase the background effect and possible isotopic fractionations in the split interface; this could also influence the hydrogen isotope values. We therefore suggest that the appropriate temperature ramp rate is an important factor in improving the precision in analyzing compound-specific hydrogen isotopes.