Glucocorticoids ameliorate neurologic symptoms in patients with glioblastoma, but their adverse effects limit long-term use. This study sought to identify factors associated with steroid taper success or failure in the early stages of glioblastoma treatment. We retrospectively reviewed steroid prescribing practices from date of surgery until one month following radiotherapy (RT) completion among 85 patients with newly diagnosed glioblastoma who were treated on a prospective clinical trial with RT and temozolomide. Sufficient information on steroid dosing was available in 72 patients included in the final analysis. The mean age was 54 years, and 65 % were men. Thirty-nine percent had a gross-total resection. Fifteen patients (21 %) tolerated steroid taper without requiring dose increase during the study. Men and patients with Karnofsky performance scale 90-100 were more likely to have a successful steroid taper. The most common symptom of taper failure was headache, but the reason for steroid increase differed among the different time intervals examined: worsening neurologic deficit in the early post-operative period, headache and non-focal symptoms during RT, and headache and seizure post-RT. Of the 50 patients in whom steroid use during RT was known, 36 (72 %) underwent dose reduction and of those, 21 (58 %) required an increase. The successful early taper of steroids in glioblastoma was associated with male gender and better functional status. Steroids are often tapered during RT, but there is frequent taper failure with this approach. A prospective trial with standardized steroid dosing regimens would be needed to verify these findings.