Spinal cord transection was induced in 3 groups of cats. The gap was surgically reconstructed using a collagen matrix bridge (Group COL), collagen matrix + pedicled omentum graft (Group COM), or gelfoam (Group GEF). After a variable observation period, animals underwent distal cord horse-radish peroxidase (HRP) injections, somatosensory evoked potentials recordings and polarographic measurement of local spinal cord blood flow (1SCBF) using the hydrogen clearance technique. The cord tissue was removed for histologic and immunohistochemical analysis. Results showed retrograde HRP labelling of proximal segmental cord neurons and somatosensory evoked potentials were present in group COM but not in COL or GEF treated animals. Local SCBF was 66% and 87% higher in COM than COL or GEF animals respectively but this increase could be reversed if flow from the pedicled omentum was clamped-off. Histologic examination of cord tissue after 45 days revealed the presence of catecholaminergic axons distal to the transection site in COM but not COL or GEF groups. Moreover, after 90 days, the rate and density of tyrosine hydroxylase immunoreactive (TH-IR) axons was 10-fold higher in COM than COL group and this was accompanied by a proportionate increase in the vascular density between the two groups. GEF treated animals showed no regeneration of transected fibers and poor blood flow pattern. These findings indicate that the placement of a pedicled omentum on a collagen matrix bridge results in near restoration of normal SCBF to the reconstructed cord region and is associated with marked regeneration of axons below the lesion site.