Background: The identification of risk factors for CNS met in stage III melanoma pts will facilitate the development of personalized surveillance strategies. Previously we characterized the incidence (3.6%, 9.6%, and 15.8% at 1, 2, and 5 yrs) and risk factors for CNS met in pts diagnosed with stage III melanoma from 1998 to 2014. It is unknown if the incidence of CNS met has changed in the ADJ IMT era. Methods: We performed an institutionally approved retrospective chart review of pts diagnosed with stage III melanoma treated with ADJ IMT between 1/1/2018 and 12/31/2020 at MD Anderson Cancer Center (MDA) or the Melanoma Institute of Australia (MIA). Exclusion criteria included mucosal or uveal melanoma or prior systemic therapy (other than ADJ interferon). Pts had to have CNS imaging within 1 month before or 4 months after stage III diagnosis. Pt and tumor features, treatments, and distant recurrence events (CNS met; non-CNS met) were captured. Time-to-CNS and time-to-non-CNS met were computed from start date of ADJ IMT to date of CNS/non-CNS met. Cumulative incidence of distant met events was determined using competing risks (death) and pts alive with no met at last follow-up were censored. Associations were determined using proportional sub-distribution hazards regression models; group differences were assessed using Gray’s test. Results: 361 pts were included in the analysis (MDA, 212; MIA, 149). With a median FU of 34.1 months (range 1.7-61.9), 128 (35.5%) of pts were diagnosed with distant mets. Among pts with distant mets, 56 had CNS met (15.5% of full cohort; 43.8% of pts with distant met); 24 (18.8% of pts with distant mets) pts had CNS met at initial diagnosis of stage IV disease. The 1-, 2-, and 5-yr rates of CNS met in the full cohort were 7%, 12%, and 17%. There was no difference in CNS met incidence between MDA and MIA. On univariate analysis, risk of CNS met was associated with AJCC stage IIID (Hazard Ratio [HR] 6.4; 95% Confidence Interval [CI] 2.2 – 18.8, p<0.001); increased primary tumor Breslow thickness (HR 1.1, CI 1.1 – 1.2, p<0.001), ulceration (HR 2.8; CI 1.5 – 5.1, p<0.001), and mitotic rate > 9/mm (HR 2.3; CI 1.1 – 4.6, p=0.025). There was a trend for decreased risk of CNS met in pts that stopped ADJ IMT due to toxicity (HR 0.4; CI 0.2 – 1.1, p=0.08). Multivariable (MV) model with these features did not identify any significant (p<0.05) associations with CNS met. Primary tumor ulceration and mitotic rate were significantly associated with non-CNS met on MV analysis. Conclusions: Despite the established impact of ADJ IMT on distant met free survival, the incidence of CNS met in stage III melanoma pts treated with ADJ IMT remains high. The results support the need for continued investigation of risk factors, prevention strategies and optimal surveillance for CNS met in stage III melanoma pts.