Type 1 diabetes (T1D) is an autoimmune disease culminating in the destruction of insulin-producing pancreatic cells. There is a need for the development of novel antigen-specific strategies to delay cell destruction, including combinatorial strategies that do not elicit systemic immunosuppression. Gamma-aminobutyric acid (GABA) is expressed by immune cells, β-cells, and gut bacteria and is immunomodulatory. Glutamic-acid decarboxylase 65 (GAD65), which catalyzes GABA from glutamate, is a T1D autoantigen. To test the efficacy of combinatorial GABA treatment with or without GAD65-immunization to dampen autoimmune responses, we enrolled recent-onset children with T1D in a one-year clinical trial (ClinicalTrials.gov NCT02002130) and examined T cell responses. We isolated peripheral blood mononuclear cells and evaluated cytokine responses following polyclonal activation and GAD65 rechallenge. Both GABA alone and GABA/GAD65-alum treatment inhibited Th1 cytokine responses over the 12-month study with both polyclonal and GAD65 restimulation. We also investigated whether patients with HLA-DR3-DQ2 and HLA-DR4-DQ8, the two highest-risk human leukocyte antigen (HLA) haplotypes in T1D, exhibited differences in response to GABA alone and GABA/GAD65-alum. HLA-DR4-DQ8 patients possessed a Th1-skewed response compared to HLA-DR3-DQ2 patients. We show that GABA and GABA/GAD65-alum present an attractive immunomodulatory treatment for children with T1D and that HLA haplotypes should be considered.