Alternatives to autografts have long been sought for use in bridging neural gaps. Many entubulation materials have been studied, although with generally disappointing results in comparison with autografts. The purpose of this study was to design a more effective neural guidance conduit, to introduce Schwann cells into the conduit, and to determine regenerative capability through it in an in vivo model. A novel, fully biodegradable polymer conduit was designed and fabricated for use in peripheral nerve repair, which approximates the macro- and microarchitecture of native peripheral nerves. It comprised a series of longitudinally aligned channels, with diameters ranging from 60 to 550 microns. The lumenal surfaces promoted the adherence of Schwann cells, whose presence is known to play a key role in nerve regeneration. This unique channel architecture increased the surface area available for Schwann cell adherence up to five-fold over that available through a simple hollow conduit. The conduit was composed of a high-molecular-weight copolymer of lactic and glycolic acids (PLGA) (MW 130,000) in an 85:15 monomer ratio. A novel foam-processing technique, employing low-pressure injection molding, was used to create highly porous conduits (approximately 90% pore volume) with continuous longitudinal channels. Using this technique, conduits were constructed containing 1, 5, 16, 45, or more longitudinally aligned channels. Prior to cellular seeding of these conduits, the foams were prewet with 50% ethanol, flushed with physiologic saline, and coated with laminin solution (10 microg/mL). A Schwann cell suspension was dynamically introduced into these processed foams at a concentration of 5 X 10(5) cells/mL, using a simple bioreactor flow loop. In vivo regeneration studies were carried out in which cell-laden five-channel polymer conduits (individual channel ID 500 microm, total conduit OD 2.3 mm) were implanted across a 7-mm gap in the rat sciatic nerve (n = 4), and midgraft axonal regeneration compared with autografts (n = 6). At 6 weeks, axonal regeneration was observed in the midconduit region of all five channels in each experimental animal. The cross-sectional area comprising axons relative to the open conduit cross sectional area (mean 26.3%, SD 10. 1%) compared favorably with autografts (mean 23.8%, SD 3.6%). Our methodology can be used to create polymer foam conduits containing longitudinally aligned channels, to introduce Schwann cells into them, and to implant them into surgically created neural defects. These conduits provide an environment permissive to axonal regeneration. Furthermore, this polymer foam-processing method and unique channeled architecture allows the introduction of neurotrophic factors into the conduit in a controlled fashion. Deposition of different factors into distinct regions within the conduit may be possible to promote more precisely guided neural regeneration.