Aging involves complex processes that impact the structure, function, and metabolism of the human brain. Declines in both structural and functional integrity along with reduced local inhibitory tone in the motor areas, as indicated by reduced γ-aminobutyric acid (GABA) levels, are often associated with compromised motor performance in elderly adults. Using multimodal neuroimaging techniques including magnetic resonance spectroscopy (MRS), diffusion magnetic resonance imaging (MRI), functional MRI as well as transcranial magnetic stimulation to assess short-interval intracortical inhibition (SICI), this study explores whether these age-related changes can be mitigated by motor learning. The investigation focused on the effects of long-term balance learning (3 months) on intracortical inhibition, metabolism, structural, and functional connectivity in the cortical sensorimotor network among an elderly cohort. We found that after 3 months of balance learning, subjects significantly improved balance performance, upregulated sensorimotor cortical GABA levels and ventral sensorimotor network functional connectivity (VSN-FC) compared to a passive control group. Furthermore, correlation analysis suggested a positive association between baseline VSN-FC and balance performance, between baseline VSN-FC and SICI, and between improvements in balance performance and upregulation in SICI in the training group, though these correlations did not survive the false discovery rate correction. These findings demonstrate that balance learning has the potential to counteract aging-related decline in functional connectivity and cortical inhibition on the “tonic” (MRS) and “functional” (SICI) level and shed new light on the close interplay between the GABAergic system, functional connectivity, and behavior.