OBJECTIVE To evaluate trans-splenic portal scintigraphy (TSPS) and per-rectal portal scintigraphy (PRPS) for diagnosis of congenital portosystemic shunts (CPSS) in dogs, and compare these results with surgical findings. METHODS Prospective, randomized cross over clinical trial. METHODS Dogs (n=42) with suspected CPSS. METHODS Dogs had TSPS and PRPS 48 hours apart; quantity of radionuclide administered was recorded. Three independent, blinded reviewers evaluated each scintigraphic study for study quality, shunt presence, number, and location of shunt termination (caudal vena cava, azygos vein). All dogs had exploratory celiotomy. Negative scintigraphic findings were confirmed with intraoperative mesenteric portography. Ameroid constrictors were placed on all extrahepatic CPSS, and hepatic biopsies were obtained. RESULTS TSPS was 100% sensitive and specific for diagnosis of CPSS and significantly (P<.05) more likely than PRPS to detect shunt number and termination. Interpretation was consistent between observers, and TSPS required significantly less radionuclide than PRPS. CONCLUSIONS TSPS was as sensitive as PRPS for detection of shunting vessels, and consistently yielded studies of higher quality, allowing detection of shunt number and location with consistent interpretation among radiologists. CONCLUSIONS TSPS provides information about shunt number and location, which allows improved surgical planning. Because it requires significantly less radionuclide, TSPS improves safety, allows for more comprehensive patient care, and earlier surgical intervention.