The aim of this study was to clarify the following questions: (1) Do newly sprouting axons use the empty neurilemmal sheaths of a flap as conduits? (2) To what extent can sensibility recover? (3) Does sensory recovery occur at the margins of the flap or at its center? (4) Does chemotaxis influence the regeneration process? In 16 free myocutaneous flaps (12 latissimus dorsi, 4 rectus abdominis) we investigated pain (pinprick), constant touch, temperature, 30- and 256-Hz vibration, and static and moving two-point discrimination between 1 1/2 and 8 years after surgery. Four flaps were anesthetic, eight recovered partly, and four had six or more modalities present all over the flap; six of ten flaps with poor recovery showed better sensitivity close to the anastomosis. Nine patients agreed to have two punch biopsies (6 mm) taken from their flaps (seven from areas with different degrees of sensory recovery and two from the center and the periphery when recovery was homogeneous). Biopsies from the areas with better sensory recovery showed slightly more nerves than the other punches in five instances, showed no difference between the two biopsies in one instance, and seemed to have more nerves in the biopsy from the area with less recovery in the last instance. The two pairs of biopsies from the homogeneously recovered flaps showed similar amounts of nerves. S-100 protein-positive neural structures (Schwann cells) were found in 13 biopsies, and neuron-specific enolase-positive nerve fibers (nerve axons) were found in 10 biopsies. Electron microscopy showed mainly unmyelinated fibers, always adjacent to vessels and sometimes with regenerative phenomena. We believe that sprouting axons probably grow into a flap attracted by chemotaxis from hair follicles, the basement membrane, and the laminin contained in the sheath. We recommend maximum scar excision at the recipient site to enhance sensory recovery in flaps. We conclude that (1) the sprouting axons primarily use the empty neurilemal sheaths as conduits, (2) sensory recovery can reach two-point discrimination equal to the donor site of the flap, and (3) sensory recovery occurs mainly from the margin of the flap but also from the bed.