Postoperative (post-op) radiosurgery (SRS) after resection of brain metastases (BM) is associated with up to 30% risk of leptomeningeal disease (LMD). Approximately 60-70% of LMD is nodular pattern in this setting, which is thought to be associated with surgical tumor spillage. We have previously demonstrated that preoperative (pre-op) SRS is a feasible alternative to post-op SRS. Herein we report patterns of failure and long terms outcomes for patients (pts) treated with pre-op SRS. The records for pts with BM treated with pre-op SRS and planned surgical resection were reviewed. Pts with classically radiosensitive tumors, prior or planned adjuvant whole brain radiotherapy (WBRT), or no cranial imaging ≥ 1 month after surgery were excluded. Pre-op SRS dose was based on lesion size and was reduced approximately 10-20% from standard. No gross tumor volume (GTV) margin expansion was used. Surgery generally followed within 48 hours. Overall survival (OS) was estimated using the Kaplan-Meier method. Intracranial event rates were estimated using cumulative incidence with competing risk of death. Radiographic pattern of LMD was classified as classical “sugarcoating” (cLMD) vs. nodular (nLMD). The study cohort consisted of 141 pts with 150 lesions treated with single fraction pre-op SRS followed by planned surgery between 2005 – 2018. Most pts had a single BM (67%), underwent gross total resection (GTR, 96%), and had non-small cell lung cancer (41%). Median interval between pre-op SRS and surgery was 2 days, median SRS dose was 15 Gy, and median GTV volume was 8.4 cc (approximately equivalent to 2.5 cm diameter lesion). The 2-year cavity local recurrence (LR) rate was 23% overall and 20% in pts status post GTR. Five of 6 pts status post subtotal resection experienced LR. Multivariate analysis demonstrated GTR and renal cell histology as significantly associated with reduced risk of LR. SRS dose, lesion volume, and time between pre-op SRS and surgery were not significant. The 2-year LMD rate was 5%. Two of the 9 patients with LMD (22%) had nLMD while 78% had cLMD. The 2-year radiation necrosis (RN), symptomatic RN, and non-LMD distant brain failure rate was 7%, 4%, and 58%, respectively. Median OS was 19 months. Thirty-seven pts with 40 lesions had long term (≥2 years of cranial imaging) follow-up. In this subset, the 5-year cavity LR, symptomatic RN, and LMD rate was 25%, 8%, and 12%, respectively. Single fraction pre-op SRS is an effective treatment for surgically resected BM with low rates of cavity LR, RN, and LMD. LMD after pre-op SRS seems to be distinct from post-op SRS with a lower overall rate and lower proportion of nLMD pattern, which supports pre-op SRS sterilization of surgical tumor spillage. Outcomes in pts with long term imaging follow-up are consistent with the overall cohort. These data support further studies of pre-op SRS, including a randomized trial of pre-op vs. post-op SRS.