Purpose <p>Taxane-induced hypersensitivity reactions (HSRs) are highly prevalent, necessitating premedication with dexamethasone; however, prolonged or high-dose exposure to dexamethasone increases the risk of adverse effects. This study evaluated whether a reduced-dose dexamethasone regimen effectively prevents HSRs while minimizing steroid-related toxicity during weekly paclitaxel chemotherapy.</p> Methods <p>This retrospective cohort study utilized electronic health records to evaluate patients receiving weekly paclitaxel. We compared a low-dose (LD) dexamethasone regimen (10 mg at the first infusion, followed by 4 mg weekly) with a conventional high-dose (HD) regimen (10 mg weekly). The primary outcome was the incidence of HSRs during the second paclitaxel infusion, analyzed using non-inferiority testing with a predefined margin of 2%. Propensity score-based inverse probability of treatment weighting was employed to control confounding, with missing data addressed via multiple imputation. Secondary outcomes included the incidence of steroid-related adverse events (AEs) such as hyperglycemia, insomnia, and serious bacterial infections, evaluated using Cox proportional hazards models adjusted for relevant clinical factors.</p> Results <p>The LD regimen was non-inferior to the HD regimen for preventing HSRs (weighted incidence, 0.72% vs. 1.18%; risk difference −0.46%; 95% CI −1.26% to 0.53%). The HD group had a significantly higher risk of hyperglycemia (adjusted HR [aHR] 1.54, 95% confidence interval [CI] 1.15 to 2.07). While insomnia (aHR 1.53, 95% CI 0.93 to 2.51) and serious bacterial infections (aHR 1.19, 95% CI 0.99 to 1.44) occurred more frequently in the HD group, these trends did not reach statistical significance.</p> Conclusion <p>A low-dose dexamethasone premedication regimen was non-inferior to the conventional regimen in preventing HSRs and was associated with a lower risk of corticosteroid-associated hyperglycemia.</p>