OBJECTIVE: There has long been interest in (and healthy skepticism of) devices and techniques to facilitate small vessel anastomoses. The purpose of this study was to assess the performance of a magnetic device to anastomose vessels. METHODS: Male canines (25 Kg) were treated daily with Plavix and aspirin. Following anesthesia, the femoral artery (2-3 mm i.d.) and vein (3-4 mm i.d.) were exposed unilaterally in three dogs and bilaterally in four dogs. An arteriotomy or venotomy of approximately 4 mm length was performed in each vessel. On each vessel, one oval magnet (Ventrica, Inc., Fremont, CA) of 0.5 mm thickness was inserted into the vessel and a second identical magnet was aligned external to the vessel, compressing the vessel wall and creating a magnetic port in the vessel. Once the magnets were firmly in place on both the artery and the vein, the two ports were approximated, completing the anastomosis. Patency was assessed after 9 weeks using a color duplex Doppler scanner. The anastomoses were explanted after 10 weeks. Hydrodynamic resistance was measured ex vivo on eight anastomoses by dividing the pressure drop across an anastomosis by the flow. RESULTS: Following placement, very high flow was evident through the anastomoses, creating visible turbulence and palpable vibration. All eleven anastomoses were patent under direct observation and palpation at implant and at explant. Ten of eleven were clearly patent on Duplex scans and one was questionable. Hydrodynamic resistance was 0.63 + 0.21 mmHg min/ml and did not vary significantly with flow. CONCLUSION: Anastomoses had 100% patency after 10 weeks and warrant further study in vessels of different sizes and locations, as well as in non-vascular structures.