In a previously documented experimental model of acute renal failure in the rat (intravenous methemoglobin-ferrocyanide) 1 ml doses of 25% mannitol (1.21–1.54 g/kg) intravenously prevented both functional and histologic renal injury within –1 h and +30 min of attempted induction. When varying mannitol doses were given 15 min following attempted induction, a sigmoid log-log relationship of plasma urea nitrogen (PUN) concentration to mannitol dose was demonstrated. Mannitol doses above 300 mg/kg (at + 15 min) are estimated to be fully protective in the model. Similarly at 15 min following attempted induction, isosmolal 1 ml doses of aqueous mannitol, mannose, sucrose, sodium sulfate, sodium chloride, THAM [tris(hydroxymethyl)aminomethane] and thriceisosmolal urea prevented acute renal failure, but isosmolal urea and dextrose, isotonic saline and 6% human serum albumin did not. Both 6% dextran (av. mol. wt. 75,000) and 10% dextran (av. mol. wt. 40,000) intensified the lesions, possibly a unique reaction in the rat. Effective agents induced higher rates of urinary flow in anesthetized, normal rats than ineffective ones. The results indicate that, in this model, (a) several osmotic diuretic agents are capable of preventing acute renal failure; (b) preventive effectiveness correlates with the ability of an agent to induce prompt, brisk diuresis; (c) the preventive therapy must be timely to be effective; and (d) the probable protective mechanism is that the agent augments the pathogenetically-significant, reduced renal tubular fluid flow. The results also imply that the protective mechanism does not depend upon extracellular or plasma volume expansion, relief of renal ischemia, osmotic loading per se, or a specific chemical or pharmacological characteristic of any single agent.