Abstract Background Non-depolarizing magnetic fields, like low frequency-pulsed electromagnetic fields (LF-PEMFs) have shown the ability to modulate living structures, principally by influencing synaptic activity and ion channels on cellular membranes. Recently, the CTU Mega 20 device was presented as a molecular accelerator, using energy up to 200 J and providing high-power (2 Tesla) pulsating fields with a water-repulsive (diamagnetic) action and tissue biostimulation. We tested the hypothesis that LF-PEMFs could modulate long-term corticospinal excitability in healthy brains by applying CTU Mega 20^®. Ten healthy subjects without known neurological and/or psychiatric diseases entered the study. A randomized double-blind sham-controlled crossover design was employed, recording TMS parameters (amplitude variation of the motor evoked potential as index of cortical excitability perturbations of the motor system) before (pre) and after (post + 0, + 15, + 30 min) a single CTU Mega 20 session on the corresponding primary right-hand motor area, using a real (magnetic field = 2 Tesla; intensity = 90 J; impulse frequency = 7 Hz; duration = 15 min) or sham device. A two-way repeated measures ANOVA with TIME (pre, post + 0, + 15, + 30 min) and TREATMENT (real vs. sham stimulation) as within-subjects factor was applied. Results A significant TIME × TREATMENT interaction was found (p < 0.001). Post hoc comparisons showed a significant effect of TIME, with significant differences at + 0, + 15 and + 30 min compared to baseline after real stimulation (all p < 0.05) but not after sham stimulation (all p < 0.05) and significant effects of TREATMENT, with significant differences at + 0, + 15 and + 30 min for real stimulation compared to sham stimulation (all p < 0.005). No significant depolarizing effects were detected throughout the (real) stimulation. Conclusions Our proof-of-concept study in healthy subjects supports the idea that non-ionizing LF-PEMFs induced by the CTU Mega 20 diamagnetic acceleration system could represent a new approach for brain neuromodulation. Further studies to optimize protocol parameters for different neurological and psychiatric conditions are warranted. Trial Registration The present work has been retrospectively registered as clinical trial on ClinicalTrials.gov NCT03537469 and publicly released on May 24, 2018