Intracerebral hemorrhage (ICH) can result in secondary brain injury due to inflammation and breakdown of the blood–brain barrier (BBB), which are closely associated with patient prognosis. The potential of the heat shock protein 90 (Hsp90) inhibitor 17-DMAG in promoting neuroprotection has been observed in certain vascular diseases. However, the precise role of 17-DMAG treatment in ICH is not yet fully understood. In this study, we found that treatment with 17-DMAG (5 mg/kg) effectively reduced hematoma expansion and resulted in improved neurological outcomes. Meanwhile, the injection of 17-DMAG had a positive effect on reducing BBB disruption in rats with ICH. This effect was achieved by increasing the levels of BBB tight junction proteins (TJPs) such as zo-1, claudin-5, and occludin. As a result, the leakage of EB extravasation, brain edema and IgG in the peri-hematoma tissue were reduced. Furthermore, the injection of 17-DMAG decreased the infiltration of neutrophils into the brain tissues surrounding the hematoma in ICH rats and also reduced the production of proinflammatory cytokines IL-6 and TNF-α. Next, we used integrative mass spectrometry (MS) and molecular docking analysis to confirm that sex determining region Y-box protein 5 (SOX5) is a potential direct target of 17-DMAG in ICH. SOX5 encodes a positive regulator of the PI3K/Akt axis, and treatment with 17-DMAG resulted in a noticeable increase in SOX5 accumulation. To further investigate the role of SOX5, we employed virus-regulated SOX5 silencing and found that suppressing SOX5 blocked the ability of 17-DMAG to suppress neutrophil trafficking. Additionally, silencing SOX5 blocked the protective effects of 17-DMAG on the BBB by inhibiting PI3K, p-Akt, and BBB TJPs levels, which led to an increase in EB and IgG leakage in the peri-hematoma tissue after ICH. Similarly, when SOX5 was knocked down, the protective effects of 17-DMAG were lost. Overall, the results of our study indicate that the injection of 17-DMAG has the potential to mitigate neuroinflammation and prevent the disruption of the BBB caused by ICH, resulting in improved neurological outcomes in rats. These positive effects are attributed to the regulation of SOX5 and activation of the PI3K/Akt pathway. These findings highlight the possibility of targeting SOX5 and the PI3K/Akt pathway as a novel therapeutic approach for ICH.