We recently reported the clinical results of a Phase I trial combining ipilimumab with a vaccine containing transgenes for prostate-specific antigen (PSA) and for a triad of costimulatory molecules (PROSTVAC) in patients with metastatic castration-resistant prostate cancer. Thirty patients were treated with escalating ipilimumab and a fixed dose of vaccine. Of 24 chemotherapy-naïve patients, 58 % had a PSA decline. Combination therapy did not exacerbate the immune-related adverse events associated with ipilimumab. Here, we present updated survival data and an evaluation of 36 immune cell subsets pre- and post-therapy. Peripheral blood mononuclear cells were collected before therapy, at 13 days and at 70 days post-initiation of therapy, and phenotyped by flow cytometry for the subsets of T cells, regulatory T cells, natural killer cells, and myeloid-derived suppressor cells. Associations between overall survival (OS) and immune cell subsets prior to treatment, and the change in a given immune cell subset 70 days post-initiation of therapy, were evaluated. The median OS was 2.63 years (1.77-3.45). There were trends toward associations for longer OS and certain immune cell subsets before immunotherapy: lower PD-1(+)Tim-3(NEG)CD4EM (P = 0.005, adjusted P = 0.010), higher PD-1(NEG)Tim-3(+)CD8 (P = 0.002, adjusted P = 0.004), and a higher number of CTLA-4(NEG) Tregs (P = 0.005, adjusted P = 0.010). We also found that an increase in Tim-3(+) natural killer cells post- versus pre-vaccination associated with longer OS (P = 0.0074, adjusted P = 0.015). These results should be considered as hypothesis generating and should be further evaluated in larger immunotherapy trials.