OBJECTIVE To optimize the surgical technique of the new extravesical seromuscular tunnel ureteroneocystostomy and to study the optimal relation between the length and width of the tunnel and the diameter of the normal and dilated ureters. METHODS The optimized technique was applied to the left ureters of 20 dogs; 10 with normal ureters and 10 after induction of ureteral dilation. The new technique was performed to maintain a ratio of 3:1 between the width of the tunnel and the diameter of the ureter and a ratio of 1:1 between the length of the tunnel and the ureteral diameter. The results were compared with those for 5 dogs in which unilateral direct ureterovesical reimplantation was performed without an antireflux procedure. All the dogs were evaluated by intravenous urography, radioisotope renography, and ascending cystography before and at 1, 3, and 6 months after ureteroneocystostomy. RESULTS All the dogs of the control group showed low-pressure vesicoureteral reflux and 3 demonstrated mild hydronephrosis and significant deterioration of the corresponding kidney. At 6-month follow-up, none of the 20 dogs subjected to the new technique showed deterioration of the function or configuration of the corresponding kidney. Moreover, the new technique was sufficient to prevent reflux in all but two dilated ureters that showed high-pressure vesicoureteral reflux. CONCLUSIONS For the extravesical seromuscular tunnel to be effective in reflux prevention without obstruction, the width of the tunnel must be triple the diameter of the ureter, and the length of the tunnel covering the intact (nonspatulated) part of the ureter should be equal to its diameter.