Background: Surgically targeted radiation therapy (STaRT), using a novel bioresorbable collagen brachytherapy device containing 4 Cesium-131 sources, is FDA-cleared for use as adjuvant radiation therapy (RT) post-resection in both newly diagnosed and recurrent intracranial neoplasms. Intraoperative initiation of brachytherapy potentially minimizes post-resection tumor regrowth with a favorable dosimetric profile compared to external beam radiation. This work sought to determine early patterns of care and the safety of this approach in recurrent GBM. Methods: This prospective multi-institutional observational study (NCT04427384) included recurrent GBM patients who underwent maximum safe resection (MSR) and permanent implantation of the device(s) (GammaTile, GT Medical Technologies, Tempe, AZ, US). Descriptive and comparative analyses regarding patient characteristics and early clinical outcomes were performed. Toxicities were categorized using the CTCAE v5.0 adverse event (AE) criteria. Results: During 10/2020–01/2023, 14 participating sites enrolled 45 patients with recurrent GBM for STaRT; 2 patients had two sites treated for a total of 47 implants. 67% of patients were treated at the first recurrence, 24% at the second, and 13% at the third, respectively. The median age was 61 (range 28-75), 36% were female, and 23% were > 65 years. 85% of patients received prior same-site RT with the median time from last RT to implantation of 14.6 months (range 3.5-57). The median maximum preoperative size was 4.2 cm (range 1.6-7.0) and the median volume was 20.8 cm³ (range 0.8-130). 68% of resections were gross-total, 17% near-total, and 15% sub-total; the median time needed for device implantation was 5 minutes (range 1.0-13). Median follow-up was 5.2 months (range 0.6-23.2). Median Karnofsky performance status (KPS) at screening, at initial postoperative assessment, and 3 months were all 80 (range 40-100). 7 attributed AEs occurred in 13% (6/45) of patients, all grade 3 AEs: 2 CSF leaks, 2 seizures, 1 cerebral edema, 1 pseudomeningocele, and 1 left hemiparesis. All except one were coded as related to both radiation and surgery; the patient with pseudomeningocele experienced a seizure at 47 days that was considered related to radiation alone. Conclusions: Early data from this prospective registry demonstrate the feasibility and safety of STaRT in recurrent GBM. Data on 6-month progression-free survival will be presented at the conference. A prospective randomized trial of adjuvant systemic therapy (AST) +MSR+STaRT versus MSR+AST is planned for initiation in 2023.