Purpose Whether an innervated reverse digital artery island flap is superior to a non‐innervated reverse digital artery island flap still remains controversial. We aimed to compare the clinical outcomes of the two flaps in repairing finger pulp soft tissue defects. Methods Medical records of patients who underwent finger pulp reconstruction between January 2007 and December 2017 were evaluated retrospectively. A total of 45 patients were included. Twenty underwent sensory nerve reconstruction with cutaneous branches of the proper digital nerve, and 25 underwent the surgery without sensory nerve reconstruction. Surgical results, complications and sensory function were collected for analysis. Sensory function was assessed by static two‐point discrimination and the modified sensory evaluation standard of British Medical Research Council. Results The average operation time of innervated flaps was 23 min longer than non‐innervated flaps. All 45 flaps survived completely. There was no significant difference in complications between groups. The average follow‐up was 22 months. At the final follow‐up, five non‐innervated flaps had no recovery of static two‐point discrimination. The average static two‐point discrimination of the remaining 20 non‐innervated flaps was larger than that of innervated flaps. Innervated flaps consistently achieved higher sensory function grades according to the modified sensory evaluation standard of British Medical Research Council. Conclusion An innervated reverse digital artery island flap can achieve better sensory function recovery in a shorter time. This procedure did not increase the incidence of complications, although it extended the operation time. It has proven to be a good technique for finger pulp reconstruction.