Isotope ratio measurements have become extremely precise in recent years, with many approaching parts-per-million (ppm) levels of precision. However, seemingly innocuous errors in signal baselines, which exist only when gas enters the instrument, might lead to significant errors. These "pressure-baseline" (PBL) offsets may have a variety of origins, such as incoherent scattering of the analyte, isobaric interferences, or electron ablation from the walls of the flight tube. They are probably present in all but ultra-high-resolution instruments, but their importance for high-precision measurements has not been investigated. We derive the governing equations for the PBL effect. We compare the oxygen triple-isotope composition of gases on three different mass spectrometers before and after applying a correction for PBLs to determine their effects. We also compare the composition of atmospheric O2 with that of several standard minerals (San-Carlos Olivine and UWG-2) on two high-precision mass spectrometers and compare those results with the differences reported in the literature. We find that PBLs lead to stretching or compression of isotopic variations. The scale distortion is non-mass-dependent, affecting the accuracy of triple-isotope covariations. The governing equations suggest that linear stretching corrections using traditional isotopic delta values (e.g., δ18 O) are rigorous for PBL-induced errors in pure gases. When the reference and sample gases are not comparable in composition or purity, however, a different correction scheme may be required. These non-mass-dependent errors are systematic and may have influenced previous measurements of triple-isotope covariations in natural materials. Accurate measurements of isotopic variations are essential to biogeochemistry and for testing theoretical models of isotope effects. PBLs are probably ubiquitous, contributing to the interlaboratory disagreements in triple-isotope compositions of materials differing greatly in δ18 O values. Moreover, they may lead to inaccurate determination of triple-isotope compositions and fractionation factors, which has implications for isotopic studies in hydrology and biogeochemistry.