The urine/plasma osmolality ratio was examined in 13 patients with acute incipient or functional renal failure, who had oliguria, a raised blood urea and a positive response to mannitol, and in 12 patients with acute organic renal failure (acute tubular necrosis) with oliguria, a high blood urea and no response to mannitol. In the former the mean ratio was 1.35 with a standard error of the mean of 0.053, and in the latter it was 1.035 with a standard error of the mean of 0.032. The difference between the two conditions was statistically significant. There was no overlap. The urine/plasma osmolality ratio was found to be an accurate test for the early diagnosis of the type of acute renal failure. It can serve as a good guide for the selection of the patients where mannitol therapy is necessary as in incipient renal failure, from those where mannitol is useless or even harmful, as in acute tubular necrosis.