A postprandial ammonia tolerance test (PPATT) was performed on normal dogs and dogs with signs that suggested they may have liver disease. All dogs underwent transcolonic scintigraphy, liver biopsy, or both and were assigned to extrahepatic disease, primary hepatocellular, and congenital portosystemic vascular anomalies (PSVA) groups. Each dog was fed a chicken and rice diet providing 25% of its estimated daily metabolizable energy requirement (MER) as an ammonia challenge. This is practical in patients with liver disease because ammonium chloride administration often causes vomiting or ammonia toxicity. Venous ammonia concentrations were measured before feeding and every 2 hours after feeding for 8 hours. No difference in mean ammonia concentrations between dogs with extrahepatic disease and control dogs was found. Therefore, the specificity of the PPATT was 100%. Dogs with hepatocellular disease showed no change in mean ammonia concentration at any time point, before or after feeding, but sensitivity was greatest when venous ammonia was measured 6 hours after feeding (sensitivity before feeding, 28%, and after feeding, 36%). Among dogs with congenital PSVA, mean ammonia concentrations were higher than the reference range at all time points before and after feeding, and peak mean ammonia concentration occurred 6 hours after feeding. In this group, the sensitivity of the PPATT was 81% before feeding and 91% 6 hours after feeding. This study demonstrates that the measurement of venous ammonia concentration is a useful test to detect congenital PSVA, and the sensitivity of the test may be improved by sampling 6 hours after feeding. The PPATT has poor sensitivity in detecting primary hepatocellular disease.