Theoretically, within the long-term marine sulfur cycle, pyrite sulfur isotopes (δ34Spy) are determined by the degree of pyrite burial that is controlled by the marine redox conditions. Thus, the secular variation of δ34Spy records the history of marine and atmosphere redox evolution. As such, large magnitude change in δ34Spy in the Proterozoic has been attributed to widespread oceanic anoxia related to generally low atmospheric pO2 levels. However, noisy δ34Spy signals in the past 250 million years cannot be explained by the frequent oscillation of seawater redox, because the global ocean was already fully oxidized. Here, by developing a numerical model, we show that δ34Spy of syndepositional pyrite is significantly affected by local factors, such as organic matter content in sediment, sedimentation rate, and seafloor redox. Thus, the noisy δ34Spy records may reflect local environmental variations rather than the oscillation of global ocean redox. Our study also suggests that δ34Spy alone cannot provide a robust constraint on the global or local ocean redox condition. We suggest that, combining with the pyrite content data and δ34Spy can help quantify local/regional biogeochemical cycles.