Richter syndrome (RS), an aggressive B cell lymphoma originating from transformation of indolent chronic lymphocytic leukemia (CLL), has been shown to be responsive to PD-1 checkpoint blockade (CPB) across clinical trials. To systematically discover the determinants of this response, we analyzed single-cell transcriptome data generated from 17 bone marrow samples longitudinally collected from 6 RS patients enrolled in a phase I study of the anti-PD1 drug nivolumab with concurrent ibrutinib (NCT 02420912). At pre-treatment sampling, RS responders (RS-R) had a larger CD8 effector/effector memory (E/EM) population than RS non-responders (RS-NR) (p=0.04, 2-sample t-test). This population was rare in normal bone marrow (N=30, p=0.001, 2-sample t-test), displayed intermediate exhaustion, preserved cytotoxicity and was marked by expression of the transcription factor and PRDM1-homolog, ZNF683. Further implicating ZNF683-expressing CD8 T cells in CPB response, comparison of gene expression between RS-Rs and RS-NRs within transcriptionally defined CD8 T cell clusters identified higher ZNF683 expression in RS-Rs. Lastly, the kinetics of ZNF683 expression correlated with response, with relative expression per cell decreasing with RS progression. Trajectory analysis supported a T cell differentiation path from GZMK+ memory to ZNF683intermediate followed by branching towards either ZNF683high or terminal exhaustion. To establish the extent to which ZNF683high T cell signatures are detectable in a broader population of RS patients, we examined bulk RNA-seq data generated from 35 independent RS biopsy specimens and found that 7 of 35 (20%) displayed a ZNF683high signature. A similar distribution of samples with ZNF683high expression, when corrected for T cell fraction, was observed in 18 of 81 (20%) CLL samples for which non-CD19 selected transcriptomic data was available and was associated with a trend towards improved overall survival (p=0.11). We also detected a similar ZNF683high signature in a large external compendium of single cell data from tumor infiltrating lymphocytes across 18 solid and 3 hematologic tumor types and in the peripheral blood of melanoma CPB responders (p<0.001 hypergeometric test). Upon evaluation of TCGA data, we detected an association with survival in melanoma cases with high T cell expression of ZNF683 (P<0.0001). To probe the cellular impact of ZNF683, a FLAG-tagged ZNF683 construct was overexpressed in Jurkat cells. ZNF683 overexpression revealed upregulation of PRF1, ITGA1, CD244, CD226 and IL2RB and downregulation of PRDM1 and IL2. CUT&RUN was performed to identify the binding sites of ZNF683, yielding twenty-four differentially regulated peaks, among which included ENCODE-identified regulatory regions adjacent to key immune genes (TNFAIP8, BTN3A2, CD69/KLRF1, ARHGAP2, CAMK4, LMO2, IL2, TCF7). Evaluation of external ATAC-seq data generated from exhausted T cell subsets revealed PD-1high tumor-infiltrating lymphocytes to have reduced accessible chromatin across identified putative ZNF683 binding sites, including TCF7. Thus, terminal exhausted cells not only lose ZNF683 expression but also display chromatin remodeling that likely abrogates its effects. By applying the tool CISTROME-GO to integrate the CUT&RUN and RNA-seq data, we observed enrichment in pathways corresponding to antigen binding and presentation, transcription factor activity, and T cell mediated cytotoxicity and cell killing upon ZNF683 expression. To determine if the association between ZNF683 expression and response to PD-1 blockade in RS could be confirmed in peripheral blood, we analyzed RNA-seq data generated from T cells isolated from pre-treatment blood samples from 7 independent RS patients treated with PD-1 therapy. We again identified ZNF683 among the top upregulated genes in RS-R (Log2 fold change = 2.13, p=0.037), along with other genes enriched in our ZNF683high CD8 E/EM signature (BATF, CORO1A, CD38, ITGB2, GZMM), while RS-NR were enriched in NK-like genes (KLRC1, FCGR3A, KLRG1, KLRF1 and FCER1G). In conclusion, we identify ZNF683 as marking a CD8 T cell population associated with CPB response in RS blood and bone marrow and provide evidence that this transcription factor regulates key cellular pathways involved in immune anti-tumor response, with implications for understanding CPB response in RS and other malignancies.