Seventeen samples of stratiform barite hosted in Middle Cambrian to middle Mississippian marine sedimentary rocks of the western Canadian Cordillera were analyzed to determine their δ34S and δ18O values. Whereas some samples had isotopic values close to those for time-equivalent evaporites, others were clearly more enriched in the heavy oxygen and sulphur isotopes. Samples with isotopic values close to the evaporite curve were from very thick stratiform barites hosted mainly in organic-rich shales, and from thin and nodular beds hosted in organic-free sedimentary strata. All samples enriched in heavy isotopes were collected from thin or nodular deposits hosted in organic-rich sedimentary strata. Enrichment of barite in heavy isotopes can be achieved by a barite precipitation–dissolution process that cycles barium between oxygenated and reduced zones in either redox-stratified wet sediment, or sea water. This cycles requires barite to either settle from oxygenated sea water into more reducing water or to precipitate within or settle into wet sediment where Eh values are decreasing in situ during sediment buildup. Because all anomalously isotopically heavy barites are hosted in organic-rich strata, barite dissolution during this cycle likely occurs through the bacterial metabolism of organic matter and sulphate, during which sulphate with 16O and 32S is preferentially broken down. The main products of this bacterial activity are CO2 and H2S, which can react to form carbonates or pyrite or can escape from the system, resulting in a depletion of light isotopes from the remaining aqueous sulphate. Because thick deposits of barite hosted in organic-rich shale, unlike smaller deposits hosted in the same strata, do not show heavy isotopic enrichment, the main factor controlling isotope fractionation is likely the rate of barite precipitation relative to barite dissolution and reduction. The δ34S versus δ18O values for six samples of Upper Devonian rocks plot fall on a line with a slope of 2. The data and observations presented here and in published reports indicate that alteration of the isotopic composition of stratiform barite strongly favours enrichment in heavy isotopes with respect to coeval sea-water sulphate. It should be possible then, to construct a barite sulphate isotopic composition age curve by averaging the lower δ18O and δ34S values for samples of the same age.