Using a rat femoral artery crush-avulsion model previously described by the authors (Rooks et al.: Microsurgery 14: 130-134, 1993), we analyzed the relative efficacy of intraarterially delivered anticoagulants against similar systemically administered intravenous anticoagulants with double blinded experimentation. The model uses a standardized crush of approximately 0.3 J and a standardized avulsion. This is followed by vascular stasis for 90 seconds after vessel repair. All rats were limited to 175 to 225 gm in weight to control vessel size. Urokinase, heparin sodium, and dextran (40,000 Dalton) were evaluated in this study. A statistically significant (p-value = 0.02) increase in urokinase efficacy was found with intraarterial delivery. (Patency rate increased from 40% to 100%). No advantage to intraarterial delivery was evident with either dextran or heparin. There was a dose related improvement in patency with heparin that was unaffected by delivery route. (Patency increased from 30% to 80% with a statistical p-value of 0.018.)