Background We assessed the efficacy and safety of the Xiaoketongbi Formula (XF) vs. pregabalin in patients with painful diabetic neuropathy (PDN). Methods Patients with PDN (n = 68) were included in a single-center, randomized, single-blind, double-dummy, parallel controlled clinical trial. The primary outcome was the change in the Brief Pain Inventory for Diabetic Peripheral Neuropathy (BPI-DPN). Secondary outcomes evaluated included the reduction of BPI-DPN >50%, changes in the numeric rating scale-11 (NRS-11) score for pain, Daily Sleep Interference Diary (DSID), Patient Global Impression of Change (PGIC), nerve conduction velocity (NCV), and adverse events. Results After 10 weeks of treatment, the BPI-DPN score reduced from 42.44 +/- 17.56 to 26.47 +/- 22.22 and from 52.03 +/- 14.30 to 37.85 +/- 17.23 in the XF and pregabalin group (P-s < 0.001), respectively. The difference in the absolute change in BPI-DPN score between both groups was -1.79 (95% CI: -9.09, 5.50; p = 0.625). In the XF and pregabalin groups, 44.1% (15/34) and 20.6% (7/34) of patients reported a BPI-DPN reduction >50% (p = 0.038), respectively. There were no significant differences between groups in NRS-11 and DSID (P-s > 0.05). A significantly greater number of patients in the XF group felt "significantly improved" or "improved" than in the pregabalin group (35.3% (12/34) vs. 11.8% (4/34), p = 0.045). The absolute change in motor nerve conduction velocity of the right median nerve was significantly different between both groups (XF group 0.7 +/- 2.3 vs. pregabalin group -2.2 +/- 4.1, p = 0.004). No serious adverse events were reported in either group. Conclusions XF is equivalent to pregabalin in reducing pain symptoms and improves the quality of life in patients with PDN. In addition, XF has the potential to improve nerve function by increasing NCV.