Currently, ERY974, a humanized IgG4 bispecific T cell-redirecting antibody recognizing glypican-3 and CD3, is in phase I clinical trials. After a first-in-human clinical trial of an anti-CD28 agonist monoclonal antibody resulting in severe life-threatening adverse events, the minimal anticipated biological effect level approach has been considered for determining the first-in-human dose of high-risk drugs. Accordingly, we aimed to determine the first-in-human dose of ERY974 using both the minimal anticipated biological effect level and no observed adverse effect level approaches. In the former, we used the 10% effective concentration value from a cytotoxicity assay using the huH-1 cell line with the highest sensitivity to ERY974 to calculate the first-in-human dose of 4.9 ng/kg, at which maximum drug concentration after 4 h of intravenous ERY974 infusion was equal to the 10% effective concentration value. To determine the no observed adverse effect level, we conducted a single-dose study in cynomolgus monkeys that were intravenously infused with ERY974 (0.1, 1, and 10 μg/kg). The lowest dose of 0.1 μg/kg was determined as the no observed adverse effect level, and the first-in-human dose of 3.2 ng/kg was calculated, considering body surface area and species difference. For the phase I clinical trial, we selected 3.0 ng/kg as a starting dose, which was lower than the first-in-human dose calculated from both the no observed adverse effect level and minimal anticipated biological effect level. Combining these two methods to determine the first-in-human dose of strong immune modulators such as T cell-redirecting antibodies would be a suitable approach from safety and efficacy perspectives.Clinical trial registration: JapicCTI-194805/NCT05022927.