The marine magnesium (Mg) cycle is the nexus bridging continent, ocean, and Earth's interior. River water is the only major Mg source, while carbonate precipitation and hydrothermal reaction are the two major processes removing seawater Mg from the ocean inventory. Reconstruction of the deep time marine Mg cycle can be approached with Mg isotopic composition of seawater (delta Mg-26(sw)), because variation of delta Mg-26(sw) may reflect the change of riverine input, carbonate precipitation or intensity of hydrothermal reaction. As the most common component of marine carbonate rocks, micrite or lime mud has been regarded as the premium material that records ancient seawater geochemical compositions and accordingly has been widely used in paleoenvironment studies. However, it is unclear whether micrite can be applied to reconstruct ancient delta Mg-26(sw). Because of the long residence time of seawater Mg, if micrite archives delta(26)Mgsw, it is predicted that delta Mg-26(micrite) should show limited temporal and spatial variation in a short (<1 million years) time scale. In this study, we measured delta Mg-26 of micrite component (delta Mg-26(micrite)) collected from Upper Paleozoic limestones in South China. Though delta Mg-26(micrite) values have limited spatial variation from the time equivalent sections, which suggests a homogeneous ancient seawater Mg isotopic composition, delta Mg-26(micrite) values show significant stratigraphic variations in all the studied sections. To further test whether such isotopic variations could reflect the change of delta Mg-26(sw), on the basis of high-resolution conodont biostratigraphic frameworks, we developed a mass balance model to quantify the marine Mg cycle. The modeling results indicate that the observed stratigraphic variations in delta Mg-26(micrite) cannot be explained by the fluctuation of delta Mg-26(sw), suggesting that extracting delta Mg-26(sw) signal from micrite is highly complicated. Such complexity might be ascribed to: (1) multiple sources of micrite (micritization of carbonate, direct precipitation from seawater/porewater, and fragmentation of calcified algae), (2) wide ranges of Mg isotopic fractionation in carbonate precipitation (e.g., the vital effect of biogenic Ca-carbonate precipitation), and (3) post-depositional overprinting at low Mg concentration in micrite. The findings of this study highlight the caution against simple interpretation of stratigraphic variability in delta Mg-26(micrite) as reflection the evolution of seawater Mg isotopic composition.