GAD-alum given into lymph nodes to type 1 diabetes patients participating in an open-label pilot trial resulted in preservation of C-peptide similar to promising results from other trials. Here, we compared the immunomodulatory effect of giving GAD-alum directly into lymph nodes versus that induced by subcutaneous administration. Samples from T1D patients (n = 6) who received 4 μg GAD-alum into lymph nodes (LNs), followed by two booster injections one month apart, and from patients (n = 6) who received two subcutaneous injections (SC) (20 μg) given one month apart were compared. GADA, IA-2A, GADA subclasses, IgE, GAD65-induced cytokines, PBMC proliferation, and T cell markers were analyzed. Lower doses of GAD-alum into LN induced higher GADA levels than SC injections and reduced proliferation and IgG1 GADA subclass, while enhancing IgG2, IgG3, and IgG4. The cytokine profile was dominated by the Th2-associated cytokine IL-13, and GAD65 stimulation induced activated CD4 T cells. Patients responding clinically best account for most of the immunological changes. In contrast, SC treatment resulted in predominant IgG1, predominant IFN-γ, higher proliferation, and activated CD4 and CD8 cells. Patients from the LN group with best metabolic outcome seemed to have common immune correlates related to the treatment. This trial is registered with DIAGNODE (NCT02352974, clinicaltrials.gov) and DIABGAD (NCT01785108, clinicaltrials.gov).