There are conflicting opinions regarding the mechanism of welding or fusion of vascular tissue by lasers. In this study, we measured the effects of saline irrigation on tissue temperature and fusion produced by argon laser welding of eight femoral and four carotid arteriovenous fistulas. Temperatures were continuously recorded using a digital thermographic camera. Forty 1-cm. welds were performed using powers of 0.50 (n = 24), 0.75 (n = 8), and 1.00 (n = 8) watt (W), with an energy fluence of 1100 J/cm2 per 1 cm segment, and cooling of the anastomotic site by saline irrigation (3 ml/minute). The "success" of fusions was determined by testing integity of the repairs by exposure to blood flow. At 0.50 and 0.75 W, successful welds were formed when the temperatures were 44.2 ± 1.6 (n = 28) and 55.0 ± 3.6°C (n = 20), with maximum temperatures of 47.9 and 59.9°C respectively. At 1 W, the tissue was desiccated and the welds disrupted when exposed to blood flow with temperatures measured at 63.7 ± 10.0°C (n = 22) and maximum of 88.0°C. Eight welds were also attempted without saline irrigation at 0.25 (n = 4) and 0.50 W (n = 4). At 0.25 W, tissue fusion was achieved but disrupted when exposed to intraluminal pressures with temperatures 50.3 ± 2.0°C (n = 10) and maximum of 52.6°C. At 0.50 W, the fusion failed after only minimal exposure to the laser energy because of tissue drying and retraction with temperatures measured at greater than 125°C. We conclude that argon laser vascular welding occurs optimally at temperatures between 44 and 60°C, that saline irrigation limits the maximum temperature and prevents drying and retraction of the tissue edges, and that welding at power levels greater than 0.75 W may have deleterious effects.