Reports indicate that L-carnitine administration before 100% lethal dose of ammonium acetate suppresses the symptoms of ammonia toxicity and prevents death in mice. However, we have been unable to confirm this observation. The cause of discrepancy between our results and the results of others was investigated with two models of hyperammonemia in mice: 1) that induced by intraperitoneal injection of urease and 2) that induced by intraperitoneal injection of ammonium acetate. L-Carnitine administration failed to protect mice against ammonia toxicity induced by intraperitoneal injection of urease. Mortality in mice treated with L-carnitine 30 min before injection of ammonium acetate was similar to that of controls pretreated with saline. Ammonia and urea levels in plasma, liver, and brain were also similar in both groups. However, the values were significantly lower than those in mice denied either pretreatment before the ammonium acetate challenge. These results indicate that pretreatment acts to reduce blood and tissue ammonia simply by diminishing the rate of absorption of the challenge, owing to the dilution of ammonium acetate upon mixing with the contents of the peritoneal cavity. Thus, any protocol that does not compare results of a putative protective agent with those obtained with an equal volume of solvents or saline runs the risk of ascribing protective property to the agent when the protection may, in fact, have been afforded by the solvent.