e21568 Background: Nodal ultrasound (US) is the preferred method for regional surveillance of clinically node-negative (cN0) primary cutaneous melanoma (CM) when sentinel lymph node biopsy (SLNB) is not performed or unsuccessful, and for pathologically node-positive (pN+) disease without completion lymph node dissection (CLND). Methods: Retrospective review was performed of nodal US surveillance from 2011-2021 in SLNB-eligible cN0 CM patients when SLNB was deferred or technically not feasible ( i.e., from failure of lymphoscintigraphic dye migration) or those with pN+ disease on SLNB, without subsequent CLND. Patients were followed for at least 2 years or until recurrence. The primary endpoints were US detection of regional nodal recurrence and comparison to clinical exam or cross-sectional imaging (PET-CT and/or CT). Results: In total, 67 patients met inclusion criteria and underwent a median of 4 nodal US exams (interquartile range, IQR 3-5) over 2-3 years of follow-up. Thirty-two (47.8%) patients declined or had unsuccessful SLNB, and 35 (52.2%) deferred CLND in favor of nodal US surveillance. Sixteen (23.9%) patients had clinical stage IA/IB (cIA/IB) CM, 15 (22.3%) had clinical stage II (cII) disease, and 36 (53.7%) had pathologic stage III (pIII) disease. Three (4.5%) patients developed satellite/in-transit metastasis, 6 (8.9%) had regional nodal recurrence, 4 (6.0%) had both local and regional recurrence, and 7 (10.4%) developed distant metastasis. Eighteen (26.9%) patients underwent tissue sampling, with 10 positive for melanoma. Nodal recurrence was observed in patients with initial cIB (1), cIIA (1), cIIB (1), cIIC (2), pIIIA (1), and pIIIC (4) disease. Three patients (cIIA, cIIB, and pIIIC due to microsatellites) had abnormal clinical exams with concurrent palpable regional adenopathy and in-transit metastasis. One patient (pIIIC) developed local satellite metastasis followed by palpable regional adenopathy. The most common surveillance method for metastatic detection was US (6/10), followed by clinical exam (3/10) and PET-CT (1/10). All metastatic nodes on US were metabolically active on subsequent PET-CT. One patient in whom PET-CT was the initial method of detection had no confirmatory US. Conclusions: Nodal US for CM requires specific radiologic expertise and is gaining traction as a cost-effective imaging modality in the United States. Our findings support the effectiveness of nodal US surveillance in cN0 or pN+ CM patients in whom SLNB or CLND is not performed.[Table: see text]