Background <p>Sentinel lymph node biopsy (SLNB) has emerged as a less morbid alternative to systematic lymphadenectomy for nodal staging in early-stage endometrial cancer. However, successful implementation requires institutional validation and an understanding of the learning curve. This study aimed to evaluate the competency curve for SLNB using cumulative sum (CUSUM) analysis.</p> Methods <p>A prospective single-centre study was conducted at a tertiary cancer institute in India from November 2023 to November 2024. Twenty-one patients with clinically apparent early-stage endometrial cancer underwent SLNB using indocyanine green (ICG) with near-infrared imaging, followed by comprehensive lymphadenectomy. The primary endpoint was bilateral SLN detection rate. Secondary endpoints included confirmed SLN retrieval and an empty packet rate. Learning curve analysis was performed using CUSUM.</p> Results <p>The overall SLN detection rate was 85.7%, with bilateral detection achieved in 52.4% of patients. Confirmed SLN retrieval was observed in 42.9% of cases. Failed mapping occurred in 14.3% of patients. The nodal positivity rate was 14.2%. The false-negative rate was 33.3% per patient among node-positive cases, with no false negatives on hemipelvic analysis. Two cases (9.5%) of empty nodal packets were identified. CUSUM analysis demonstrated significant improvement in bilateral SLN detection and confirmed retrieval after the 9th case (<i>p</i> = 0.0014), with a plateau observed after approximately 18 cases. Each additional case increased the odds of confirmed SLN retrieval by 7% and reduced the odds of empty packet occurrence by 3%.</p> Conclusions <p>SLNB in endometrial cancer demonstrates a clear learning curve, with significant improvement after 9 cases and near-competency achieved by 18 cases. Empty packet rate may serve as an additional quality indicator. With structured training and adherence to standardized protocols, SLNB is feasible even in resource-limited settings.</p>