Distinct, non-mass-dependent 17 O depletion was reported in barite from Marinoan cap dolostones (~635Ma) in South China, and has been interpreted as an indication of an extremely high-pCO 2 atmosphere (Bao et al., 2008). There is a significant variation in the magnitude of the 17 O anomaly even among samples from the same outcrops. Understanding the origins of this barite and particularly the source of sulfate is critical to interpreting the anomalous 17 O signature and its implication on sulfur and oxygen cycles at the aftermath of Marinoan glacial meltdown. In this study, we analyzed 160 micro-sampled barite samples from 20 hand specimens collected from two cap dolostone sections in South China. Together with field, petrographic, and Sr isotope data, the expanded dataset 1) confirms the large variability in ∆ 17 O (from −0.09‰ to −0.87‰, VSMOW), δ 18 O (+13.3‰ to +23.5‰, VSMOW), and δ 34 S (+20.0‰ to +45.5‰, VCDT) of barite; 2) demonstrates a hyperbolic relationship between the ∆ 17 O and δ 34 S of various barite samples with more negative Δ 17 O values generally corresponding to lower δ 34 S values; 3) reveals that individual barite crystal fans and fans of the same layer possess a well-clustered set of δ 18 O, Δ 17 O, and δ 34 S values; and 4) shows, correspondingly, that barite crystal fans of different layers bear different sets of Δ 17 O, δ 18 O, and δ 34 S values. The study suggests that 17 O-depleted barite crystals were formed under supersaturation when Ba 2+ from sulfate-free deepwater came to mix with sulfate-bearing shallow water. There were multiple upwelling episodes which resulted in multiple barite layers. The large variability in sulfur and triple-oxygen isotope composition and the high 87 Sr/ 86 Sr ratios indicate that the two sites from South China were sufficiently close to the continent so that the isotopic composition of sulfate was easily influenced by changes in riverine flux when seawater sulfate concentration at that time was exceedingly low. Our depositional model for the barite is consistent with the proposed sequence of events after the meltdown of Marinoan glaciation in South China by Zhou et al. (2010), and is also consistent with the proposed atmospheric origin of sulfate 17 O depletion by Bao et al. (2008).