Background: Various factors have been reported to affect the obliteration of brain arteriovenous malformations (AVM) following stereotactic radiosurgery (SRS). This meta-analysis was conducted to identify the factors potentially associated with AVM obliteration after SRS. Methods: We comprehensively searched databases and included studies that evaluated predictors of AVM obliteration after SRS using Cox proportional hazard regression analysis. Hazard ratios (HRs) with 95% confidence intervals (CIs) were utilized as effect estimates. Results: Twelve studies, involving 4415 AVM cases, were included. According to combined estimates on univariate (UV) and multivariate (MV) analysis, age, gender and prior haemorrhage did not affect the closure probability. The following factors showed a significant and independent association with increased AVM obliteration: smaller AVMs maximal diameter (MV, HR: 1.32), smaller AVM volume (MV, HR: 1.05), AVM volume <10-15 cc (MV, HR: 1.55), higher margin dose (MV, HR: 1.05), margin dose ≥17-18 Gy (MV, HR: 3.71) and early treatment period (MV, HR: 1.78). Previous embolization and deep-seated AVM were independently negative predictors of obliteration whereas deep venous drainage was a positive predictor. Prior surgery, compactness of nidus, lower grading scores and higher SRS maximum dose were associated factors subject to confounding factors. Conclusion: Multiple factors associated with obliteration should be taken into consideration for selection of candidates with AVMs for SRS. AVM volume and radiation dose are the most prominent factor in assessing obliteration after SRS. Age, gender and prior haemorrhage may not affect the consideration of SRS treatment. Cautious use of SRS is needed for previously embolized AVM patients. * Two sets of data were extracted for presentation, namely data for AVMs treated by GK alone (81 patients) and data for AVMs with post-GK embolization (17 patients). † Two sets of data in different treatment periods were extracted.