We have proposed that it is sufficient to provide a simple substratum on which regenerating axons may traverse a gap in a peripheral nerve. To test this hypothesis we set up a new experimental model in which sutures were used to bridge a 10 mm long defect in a peripheral nerve. A defect was created bilaterally in 25 rat sciatic nerves. The cut ends of the nerve were joined by three laps of a continuous suture, on one side with 8/0 polyamide (non-absorbable) and on the other with 8/0 polyglactin (absorbable), leaving a 7 mm gap. At two weeks a matrix that contained capillaries, fibroblast-like cells, and mononuclear cells had formed between the nerve endings, and the sutures were surrounded by foreign-body-like tissue reactions. At four weeks axons had grown into the distal nerve segment on both sides in 65%-90% of the cases as indicated by a response to the pinch reflex test and immunocytochemistry for presence of neurofilament protein. Axons were organised in minifascicles and these tended to grow larger as the demarcation of the entire regenerated segment by a perineurial-like structure improved with time. At 12 weeks axonal counts of cross-sections of the distal tibial trunk showed many myelinated nerve fibres but no significant difference in axonal counts or degree of myelination between the polyamide and polyglactin sides. The results show that conventional sutures alone are sufficient to support regeneration across a short gap in a peripheral nerve, a method that may be of potential clinical value.