After decades of applying a one-setup-fits-all approach to transcranial magnetic stimulation (TMS), increasing evidence of substantial intra- and inter-subject variability urged a move away from this generalized TMS approach toward subject-specific stimulation. This new concept started with synchronizing TMS to the characteristics of the ongoing electroencephalography (EEG) signal instead of ignoring the ongoing brain state at the time of stimulation. This approach has been termed brain state-dependent stimulation. Targeting predetermined specific brain states, such as a specific phase of an ongoing EEG oscillation, produced the desired effect at the population level, e.g., a long-term potentiation-like increase in excitability of the stimulated brain region. To advance to closed-loop TMS, the optimal stimulation parameters, including the targeted brain state, would be determined based on real-time analysis of feedback during stimulation, as opposed to open-loop TMS, which does not take this feedback into account. This chapter presents the evolution of TMS toward personalized approaches and discusses the milestones that have been achieved and the obstacles that remain to be overcome.