In the pase decade, several experimental models of acute renal failure (ARF) have been evaluated with micropuncture and hemodynamic techniques. Five of these models have been most extensively studied: glycerol injection, renal artery clamping, intrarenal norepinephrine infusion, uranyl nitrate, and mercuric chloride administration. In the first three models, renal ischemia is the initiating insult, whereas in the two nephrotoxic models a direct effect of the agent on cellular integrity is also seemingly operative. In all of these models, renal blood flow 24--48 h after the initial insult either spontaneously returns to normal or can be elevated to this level with volume expansion but without restoration of the glomerular filtration rate. Therefore, the maintenance of ARF in these various models is due to other factors, which include tubular obstruction, leakage of filtrate across damaged tubular epithelium, and a decrease in the glomerular capillary ultrafiltration coefficient. In a given model, one or all three of these alterations may be present. Although these various models may not be completely analogous to the clinical setting, they have provided powerful tools for the study of ARF and their use has greatly increased our knowledge in this field.