Identify prognostic factors for progression-free survival (PFS) and overall survival (OS) after reirradiation (re-RT) for recurrent high grade glioma. An institutional database was queried for patients with high grade glioma (HGG) who received re-RT for progressive disease from 2010 to present. PFS and OS after re-RT were estimated using the Kaplan-Meier method, and prognostic variables were examined using univariate and multivariate Cox models. Receiver operative curve (ROC) analysis was used to determine best predictive thresholds for continuous variables. Fifty-eight eligible patients had received surgery and adjuvant radiation ± temozolomide for initial diagnosis of HGG (51 grade IV, 7 grade III). The median time to first progression after initial radiation was 11 months. Prior salvage therapy before re-RT included chemotherapy (60%) and surgery (45%). The median number of separate progression events before re-RT was 1 (range 0 – 5). The median time from first progression to re-RT was 2.7 months, and from initial radiation to re-RT was 18 months. 36% received single fraction stereotactic re-RT (SRS) (median 18 Gy, range 14 – 19 Gy) and 64% received fractionated re-RT (median 35 Gy in 10 fractions, range 10 – 60 Gy in 2 – 30 fractions). The median biologically effective dose (BED10) of re-RT was 47 Gy (range 15 – 72). The median planning target volume (PTV) was 16.8 mL (range 0.4 – 783.6). 50% received concurrent chemotherapy and 36% received bevacizumab concurrent and/or adjuvant to re-RT. Acute (≤ 3 months) toxicity ≥ grade 3 was 7%. The median PFS after re-RT was 4.7 months (1 year PFS 18%), and the median OS was 11 months (2 year OS 21%). By univariate analysis, lower PFS was significantly (p<0.05) associated with shorter time to first progression after initial radiation, lower KPS, and lower re-RT dose (BED10). Lower OS was associated with shorter time to first progression after initial radiation, lower KPS, and larger PTV. Time from initial RT to re-RT, time from first progression to re-RT, use of SRS, and chemotherapy or bevacizumab at re-RT were not significantly associated with PFS or OS. ROC analysis of time to first progression and re-RT dose showed best predictive thresholds at time > 12 months and BED10 > 42 Gy. All significant univariate factors retained significance on multivariate analysis except re-RT dose (p = 0.07). Reirradiation was well tolerated with infrequent high grade acute toxicity. PFS and OS after re-RT for high grade glioma were both predicted by time to progression after initial radiation. Published prognostic scores have used total time from first to second radiation courses; however in our series the period from initial progression to re-RT did not add prognostic information. There was evidence for a dose threshold of BED10 > 42 Gy (> 16 Gy in 1 fraction, 27.5 Gy in 5 fractions, or 32 Gy in 10 fractions) irrespective of radiotherapy technique. Use of chemotherapy and bevacizumab with re-RT were not associated with improved PFS or OS.