This review summarizes the current understanding of the renal biopsy in "acute tubular necrosis" and the attempts to mimic this phenomenon in animal models. Paradoxically, only very limited necrosis is present in the biopsy of patients with this condition and differences in biopsies of patients with sustained and recovering renal failure cannot be clearly defined. The small amount of material examined, the variation in timing of the biopsy, the ability of the nephron to recover from sublethal injury, and the complexity of the clinical situation compound the difficulties in understanding this condition. Morphological findings in the animal studies are not equivalent to those in the human biopsy of "acute tubular necrosis," because they either have too much proximal tubular necrosis (ischemia-reflow model) or show severe injury to distal nephron segments (distal nephron model), the degree of which has not been clearly documented, as yet, in human material. The direct relevance of animal models in part may be tested by new noninvasive methods that define and quantify excreted proteins that reflect nephron injury or measure the status of renal oxygenation by radiological imaging techniques. Finally, it may be time to re-examine the morphology of "acute tubular necrosis," utilizing new techniques that illustrate induction of heat shock proteins, sublethal and apoptotic cellular injury, and alteration of gene expression.