The treatment of melanoma with checkpoint inhibitor (CPI) therapy has produced rapid and substantial gains in survival, and as a result, the indications for its use have steadily increased. As CPIs are used in more patients, and particularly as part of a neoadjuvant therapeutic strategy, the significance of the challenge presented by toxicities has also increased. Immune-related adverse events (irAEs), always a meaningful concern in the use of CPIs, can present with increased severity in the neoadjuvant setting, and can delay timely progression to surgery in addition to their impacts on morbidity, mortality, and the ability to maintain an uninterrupted course of CPI treatment. irAE in the skin, the most frequently affected organ, can present with a variety of morphologies that mimic a collection of naturally-occurring inflammatory dermatoses. The management strategies of these irAE is built on the treatment of their naturally-occurring namesakes. Much about the immunologic mechanisms of the irAE remains poorly understood, and as a result, management decisions are made empirically, to mixed results. Here, we propose to address this challenge by leveraging the distinct strengths of our team’s institutions and laboratories to compile a thorough characterization of the immune infiltrates across morphologies of cutaneous irAE, and compare these to the corresponding naturally-occurring dermatoses. Beginning with a substantial existing repository of patient samples, we will use sequential-slice tissue sections to generate paired single-cell and spatial transcriptomic analyses of both irAE and morphologically-matched dermatoses occurring outside the context of CPIs. This strategy will allow us to combine the power of single-cell resolution to generate information about population shifts and transcriptional program activation with the tissue-level context and cell-cell interaction mapping that spatial sequencing makes possible, generating in the process a first-of-its-kind library of transcriptional data across a class of immune-mediated skin disease. We will then interrogate this library to reveal features that connect irAE to (or distinguish them from) their naturally-occurring counterparts, along with those features that define cutaneous irAE as a class of toxicities to melanoma therapeutics. Building on these results, we will utilize multiplex RNA in situ hybridization, quantitative RT-PCR, and protein-level cytokine arrays to validate significant driver genes, while assessing the relationship between in-skin inflammatory response and patient-level shifts in circulating cytokine levels. Collectively, this work will represent a foundational step toward the understanding of the breakdown of selftolerance that is commonly a feature of the treatment of melanoma patients with CPI, and will provide mechanistic support for what are currently empiric treatment decisions that can determine the success of multiple CPI delivery strategies.