Composite biomaterials designed for the repair of abdominal wall defects are usually composed of a reticular component and a second component, or barrier, which is laminar (absorbable or non-reabsorbable) and which is placed in contact with the visceral peritoneum. This study was designed to evaluate how the composition of this second component affects the biological behavior of the biomaterial and the formation of adhesions.Defects (7 x 5 cm) comprising all the tissue planes except the skin were created in the anterior abdominal wall of white, New Zealand rabbits. In group A (n = 12) the defects were then repaired with composite prostheses with a non-reabsorbable physical barrier, polypropylene + ePTFE (PL + ePTFE) and PL + polyurethane (PL + PU). In group B (n = 12) the defects were repaired with a reabsorbable chemical barrier, polyester + a polyethyleneglycol/glycerol film (PO + gl) and PL + hyaluronate (PL + hy). Fourteen days after surgery, the animals were sacrificed and specimens were taken for light and scanning electron microscopy and immunohistochemical labeling for macrophages (RAM-11). Adhesions forming at the prostheses/visceral peritoneum interface were quantified.All the materials showed optimal tissue infiltration and mesothelialization. Adhesion formation was similar in prostheses with a physical barrier (PL + ePTFE, 0.49 +/- 0.14%; PL-PU, 0.29 +/- 0.2%). In contrast, those with a chemical barrier showed a significant difference in adhesion formation (PO + gl, 0.55 +/- 0.06%; PL + hy, 18.55 +/- 4.96%; p < 0.05).Physical barriers seem to induce similar adhesions, while adhesions formed to prostheses with chemical barriers can vary considerably, possibly depending on the chemical composition of the barrier.