Abstract: Zirconium is a key fission product in high-level liquid waste, originating from the neutron-induced fission of uranium and plutonium in nuclear fuel, and its content in the waste stream can be substantial depending on burnup and cooling time. Its solubility in borosilicate glass directly governs the achievable waste loading and the long-term chemical durability of the solidified waste form, because excess ZrO₂ tends to precipitate as refractory phases that compromise the glass network integrity. In this study, a systematic experimental investigation was conducted to determine the solubility thresholds of ZrO₂ in a typical borosilicate glass system by precisely controlling two variables: the total simulated waste content(representing a multi-component oxide mixture) and the added ZrO₂ amount. Two distinct addition modes were examined in parallel. The first mode, “tolerance capacity”, involved substituting ZrO₂ for an equivalent mass of the simulated waste mixture, thereby assessing the maximum tolerable ZrO₂ level while keeping the total waste loading constant. The second mode, “co-solubility”, involved adding ZrO₂ simultaneously with the full simulated waste composition to evaluate the mutual solubility limits under realistic multi-component conditions, where various cations compete for network-modifying sites. Waste loadings were set at 18%, 20%, and 22%(mass fraction), and ZrO₂ content was varied across a wide range to bracket the expected thresholds. Characterization of the glass waste forms was performed using scanning electron microscopy(SEM) for microstructure observation, energy-dispersive spectroscopy(EDS) for elemental mapping to detect Zr-rich segregations, and X-ray diffraction(XRD) for unambiguous identification of crystalline phases. The results reveal a clear dependence of the ZrO₂ solubility limit on both waste loading and addition mode. Under tolerance capacity, at waste loadings of 18% and 20%, the solubility threshold remains stable at 6%-8%(mass fraction) ZrO₂, indicating sufficient free volume and modifier cations to incorporate Zr into the silicate network. However, when the loading increases to 22%, the threshold drops significantly to 4%-6%, because the available network modifier sites become increasingly occupied by other waste components such as rare earths and transition metals, reducing the accommodation capacity for Zr. Under co-solubility, with simulated waste content ranging from 18% to 22%(mass fraction), the threshold is consistently 4%-6%, indicating that simultaneous presence of all waste constituents imposes a more restrictive limit due to competitive cation interactions and the saturation of the glass network. Importantly, exceeding the respective threshold leads to pronounced phase separation, with crystallization of ZrSiO₄ and residual ZrO₂ phases, as confirmed by XRD. This transformation from a homogeneous glassy state to a heterogeneous composite not only impairs the chemical durability—increasing the leaching rate of radionuclides—but also reduces the mechanical strength and thermal stability of the waste form. The present work establishes quantitative solubility boundaries for ZrO₂ under realistic conditions, provides mechanistic insights into the competition between glass network incorporation and phase precipitation, and offers critical data for optimizing ZrO₂ loading and adjusting glass formulations in practical vitrification processes. These findings are of significant engineering value for improving waste volume reduction, minimizing secondary waste, and ensuring long-term safety in nuclear waste management, thereby supporting the design of more robust and efficient glass waste forms for geological disposal.