Introduction Transcranial direct current stimulation (tDCS) improves postural response to perturbation in patients with Parkinson's disease (PwPD). However, the influence of baseline characteristics such as clinical/cognitive and postural performance on the response to tDCS remains unclear. Objective To investigate whether baseline level of postural control (performance during sham condition) and clinical/cognitive characteristics are associated with tDCS-related changes in postural responses to external perturbations in PwPD. Methods Twenty-four PwPD participated in this study. Clinical assessment included disease severity, disease duration, levodopa equivalent dose and global cognition. Anodal tDCS protocols targeting the primary motor cortex were applied in two separate sessions (at least 2 weeks apart): active (2 mA for 20 min) and sham stimulation. Seven trials with the backward translation of the support base (20 cm/s and 5 cm) were performed after tDCS. Postural outcomes included the recovery time to stable position and onset latency of the medial gastrocnemius (MG). Pearson and Spearman correlation tests were performed. Results No significant correlations were observed between clinical/cognitive characteristics and tDCS-related changes in postural responses. Negative associations were observed between the baseline level of postural control and tDCS-related changes in postural responses for the recovery time (r = −0.657; p < 0.001) and the MG onset latency (rs = −0.539; p = 0.007). PwPD with worse baseline postural control demonstrated greater improvement after active stimulation. Conclusions Findings suggest that tDCS-related effects on postural response to perturbation are related to the baseline level of postural control, but not to clinical characteristics in PwPD. Those with worse baseline postural control responded better to tDCS.