Chimeric antigen receptor (CAR) T-cell therapy has fundamentally changed the therapeutic landscape for haematological malignancies. CAR T-cell therapy involves the collection of a patient's T cells, ex-vivo genetic modification of the cells to encode a synthetic receptor that binds a specific tumour antigen, and then re-infusion of these cells back into the patient. The clinical success of CAR T-cell therapy in blood cancers has generated enthusiasm for testing the technology in solid tumours. However, the biology of solid tumours is more complex than that of haematological malignancies. Unfortunately, CAR T-cell therapy for solid tumours suffered a setback on June 2, 2021, when Tmunity Therapeutics stopped its phase 1 clinical trial in men with prostate cancer because two patients died from immune effector cell-associated neurotoxicity syndrome (ICANS). Tmunity Therapeutics had been developing prostate-specific membrane antigen (PSMA)-directed, TGF-β-insensitive CAR T cells. Prostate cancer cells secrete TGF-β as an immunosuppressive factor, so the investigators hypothesised that using a PSMA-directed CAR T cell with coexpression of a dominant-negative TGF-β receptor would enhance antitumour immunity in patients with prostate cancer. Early trial results had shown no dose-limiting toxicities, and only one, reversible, case of cytokine release syndrome (CRS). ICANS, however, although seen less frequently than CRS, has been reported in up to half of all patients receiving CAR T-cell therapy for various haematological cancers and thus was also a risk in solid tumour trials. ICANS needs prompt diagnosis and appropriate management to avoid rapid clinical deterioration. Unlike CRS, the precise underlying pathophysiology of ICANS remains poorly understood, and although the acute symptoms of CRS and ICANS are generally reversible with supportive care, steroids, and immunosuppressive drugs, they can be associated with substantial morbidity, with some patients requiring admission to an intensive care unit. In addition to CRS and ICANS, the development of CAR T-cell therapy for solid tumours faces many fundamental challenges, the most difficult being the identification of an ideal target antigen. Haematological cancers typically express a single, specific, tumour-associated antigen, whereas solid tumours have substantial antigen heterogeneity. Of concern, many antigens associated with solid tumours are also expressed on healthy tissue, raising the risk of off-target adverse events. Genetic modification technologies can help to circumvent this risk, but they are not a failsafe solution. Other major challenges include how to traffic the CAR T cells to the tumour bed because of a lack of appropriate cytokines in the tumour milieu; how to prevent CAR T cells from losing their effectiveness over time (so-called T-cell fitness); how to lymphodeplete the tumour to allow for CAR T-cell expansion in the absence of effective lymphodepleting chemotherapies; and, finally, how to overcome the hostile immunosuppressive conditions of the tumour microenvironment. A recent development might help to overcome two of these challenges. Using a new CAR T-cell technology called synthetic Notch (or synNotch), CAR T cells can be made to target specific cancers very precisely. Researchers at the Children's Hospital Los Angeles (Los Angeles, CA, USA) have developed a modified version of a CAR T cell for treating neuroblastoma. The synNotch protein on the surface of the T cell is designed to recognise an antigen called GD2. When it does, the synNotch protein then instructs the T cell to activate its CAR T-cell properties, enabling it to recognise a second antigen called B7H3. This gating property is key to minimising toxicity, as healthy cells sometimes express one of these antigens, but rarely both. Neuroblastoma, however, has both GD2 and B7H3 antigens. Interestingly, synNotch CAR T cells are metabolically more stable than regular CAR T cells, enabling them to target a tumour over a longer timeframe and overcome the problems associated with CAR T-cell fitness. The use of CAR T-cell therapy has yielded excellent early-to-mid-term results in patients with haematological cancers, but the transfer of this technology to solid cancers faces more substantive biological barriers and risks to patients. Despite many ongoing clinical trials, very few have reported any results and all are at early stages. Consequently, many unknowns remain. The potential is evident, but it will only be realised through carefully designed trials incorporating thorough patient monitoring and stringent stopping rules coupled to detailed preclinical translational research.