Objective We investigated if anodal transcranial direct current stimulation (A-tDCS), applied over the supplementary motor areas (SMAs), could improve gait initiation in Parkinson's disease (PD) with freezing of gait (FOG). Methods In this double-blinded cross-over pilot study, ten PD with FOG underwent two stimulation sessions: A-tDCS (1 mA, 10 min) and sham stimulation. Eight blocks of gait initiation were collected per session: (1) pre-tDCS, with acoustic cueing; (2) pre-tDCS, self-initiated (no cue); and (3-8) post-tDCS, self-initiated. Gait initiation kinetics were analyzed with two-way repeated measures ANOVAs for the effects of A-tDCS. Results A-tDCS did not significantly improve the magnitude or timing of anticipatory postural adjustments or the execution of the first step during self-initiated gait compared with baseline measures (p > .13). The lack of significant change was not due to an inability to generate functional APAs since external cueing markedly improved gait initiation (p < .01). Conclusions A single dose of A-tDCS over the SMAs did not improve self-initiated gait in PD and FOG. Alternative approaches using a different dose or cortical target are worthy of exploration since individuals demonstrated the capacity to improve. Significance Neuromodulation strategies tailored to facilitate SMA activity may be ineffective for the treatment of gait initiation impairment in people with PD and FOG.