The compliance mismatch hypothesis as a cause of failure of small diameter grafts was investigated using 4 mm internal diameter, 5 cm long Biolized polyurethane grafts. The luminal surface of 0.5 mm thick Biomer sponge grafts was covered with 100 microns of crosslinked gelatin. Compliance matched, and noncompliant grafts had in vivo compliance values of 13.5 +/- 2.89 and 1.0 +/- 0.7 X 10(-2%) per mmHg, respectively. In vivo compliance defined as 2 X dD/dP X 1/Do was measured in situ using the ultrasonic Hokanson probe. Compliance values of compliant grafts equaled that of the canine carotid artery at 100 mmHg mean arterial pressure. Grafts were implanted using an end-to-end anastomosis in 14 dogs for 6 wk. Patency rates for compliant vs. noncompliant grafts were 64% (9/14) and 50% (7/14). Compliance values for compliant grafts decreased to a mean of 7.14 X 10(-2%) per mmHg. Anastomotic intimal hyperplasia at the proximal and distal ends of compliant and noncompliant grafts measured 239.9 +/- 128, 197 +/- 129 microns, 338.9 +/- 273.6, and 304.3 +/- 179.3 microns, respectively. There were no significant differences in localized anastomotic intimal hyperplasia at either graft end for either graft type. This study did not show any positive effects of graft compliance.