The sentinel lymph node (SLN) is the first lymph node encountered by a metastatic cancer cell and serves as a predictor of poor prognosis, as cases with clinically occult SLN metastases are classified as stage III with elevated rates of recurrence and diminished overall survival. However, the dynamics of immune infiltrates in SLNs remain poorly characterized. Here, using an unbiased cellular indexing of transcriptomes and epitopes by sequencing technique, we profiled 97,777 cells from SLN tissues obtained from patients with stages I/II and III cutaneous melanoma. We described the transcriptional programs of a multitude of T, B, and myeloid cell subtypes in SLNs. Based on the proportions of cell types, we determined that SLN subtypes stratified along a naive → activated axis; patients with a "high activated" signature score appeared to be undergoing a robust melanoma antigen-driven adaptive immune response and, thus, could be responsive to immunotherapy. Additionally, we identified transcriptomic signatures of SLN-infiltrating dendritic cell subsets that compromise antitumor immune responses. Our analyses provide valuable insights into tumor-driven immune changes in the SLN tissue, offering a powerful tool for the informed design of immune therapies for patients with high-risk melanoma.