BACKGROUND: Glioblastoma is a malignant brain tumor with poor outcomes. Chimeric antigen receptor-T (CAR-T)-cell therapy is a possible new intervention in solid tumors using T cells engineered with cancer-specific antigens. This approach is challenged by limited T-cell trafficking to solid tumors, the immunosuppressive tumor microenvironment, and glioblastoma-specific uncertainties. Several first-in-human trials have now trialed CAR-T therapy for glioblastoma. We undertake a systematic review of these Phase I trials to draw early lessons about the safety and feasibility of this approach. METHODS: Systematic review of all published clinical trials using CAR-T therapy for glioblastoma on July 31, 2024, from 5 databases. RESULTS: Thirteen published studies of Phase I trials of CAR-T therapy for glioblastoma ( = 128 patients). Six molecular targets were used, most commonly EGFR family (7 studies) and IL13a2 (4 studies). There were 141 severe adverse effects (SAEs) and 2 dose-limiting toxicities. SAEs were most commonly neurological or hematological, and most commonly observed with doses over 1 × 10 cells. Routes of delivery included intravenous, intraventricular, intracavitary, and intratumoral. Several participants across trials demonstrated transient responses, but efficacy across trials was difficult to compare, given heterogeneous reporting of outcomes. CONCLUSIONS: Several CAR-T strategies have now been trialed preliminarily for glioblastoma. There appears to be a signal of efficacy, with 56 of 128 reported patients demonstrating at least some measure of response. Central delivery of CAR-T cells appears safe with doses up to 2.5 × 10 cells well-tolerated. Subsequent CAR-T trials should standardize reporting of outcomes for better comparison across trials.