Background: Preoperative single fraction radiosurgery (SRS) and postoperative fractionated SRS delivered over 3–5 fractions have been shown to have favorable outcomes compared to postoperative single fraction SRS for resected brain metastases. No study has directly compared these 2 treatment approaches. Methods: Records for patients with resected brain metastases treated with either single fraction preoperative SRS or fractionated (3-5 fractions) postoperative SRS were reviewed. Preoperative SRS was 10-20% dose reduced compared to standard and surgery generally followed within 48 hours. Eligibility criteria included solid tumor metastases, 1 brain metastasis resected, and no previous cranial RT. Fine-Gray and Cox multivariable (MVA) and propensity score matched (PSM) analyses were used. Results: A total of 330 patients (137 preoperative; 193 postoperative) were included. Median dose was 15 Gy in 1 fraction and 24 Gy in 3 fractions, respectively. In MVA, preoperative SRS was significantly associated with higher risk of cavity local recurrence (LR, hazard ratio (HR) 2.04, p=0.002) and lower risk of leptomeningeal disease (LMD, HR 0.41, p=0.05). There was no difference in adverse radiation effect (ARE) or overall survival (OS) between groups. In the PSM analysis (65 matched pairs), 1-year outcomes for preoperative vs. postoperative SRS were as follows - cavity LR: 22.9% vs. 3.1%, p< 0.001, LMD: 4.2% vs. 15.5%, p=0.04, ARE: 3.2% vs. 7.9%, p=0.73, composite endpoint (cavity LR, symptomatic ARE, or LMD): 27.6% vs. 20.1%, p=0.33, OS: 56.3% vs. 62.3%, p=0.8. Conclusions: Preoperative single fraction SRS and postoperative fractionated (3-5 fractions) SRS demonstrate distinct patterns of failure. Compared to postoperative SRS, preoperative SRS was associated with increased risk of cavity LR and lower risk of LMD in both multivariable and PSM analyses. There was no difference in risk of ARE or OS. Methods to reduce preoperative SRS cavity LR, such as with higher dose fractionated regimens, should be considered.