Single-pulse and repetitive transcranial magnetic stimulation (rTMS) are used in clinical practice for diagnostic and therapeutic purposes. However, rTMS-based therapies that lead to a significant and sustained reduction in neuropsychiatric symptoms remain scarce. While it is generally accepted that the stimulation frequency plays a crucial role in producing the therapeutic effects of rTMS, less attention has been dedicated to determining the role of the electric field strength. Conventional threshold-based intensity selection approaches, such as the resting motor threshold, produce variable stimulation intensities and electric fields across participants and cortical regions. Insufficient standardization of electric field strength may contribute to the variability of rTMS effects and thus therapeutic success. Computational approaches that can prospectively optimize the electric field and standardize it across patients and cortical targets may overcome some of these limitations. Here, we discuss these approaches and propose that electric field standardization will be instrumental for translational science frameworks (e.g., multiscale modeling and basic science approaches) aimed at deciphering the subcellular, cellular, and network mechanisms of rTMS. Advances in understanding these mechanisms will be important for optimizing rTMS-based therapies in psychiatry.