Halothane-induced changes in renal function have generally been attributed to alterations in systemic hemodynamics, sympathetic tone, and various hormones. Studies were performed to determine whether halothane directly affects the kidney. Twenty-one canine kidneys were perfused in vitro utilizing hemodilution, pulsatile flow, and membrane oxygenation. Temperature and arterial blood-gas variables were controlled and mean and pulse pressures were maintained. Four experimental periods (I-IV) (each consisting of two 10-min sample collection periods) were conducted, with a 20-min “rest” period between succeeding experimental periods (elapsed time = 140 min). Responsiveness was assured by obtaining a normal response to furosemide, acetylcholine, or epinephrine after Period IV. In eight additional kidney preparations halothane was administered to achieve either a “low” (17 + 3 mg/100 ml) or “high” (35 ± 5 mg/100 ml) concentration in Period II, the sequence reversed for Period III, and halothane eliminated by Period IV. Halothane produced marked increases in blood flow (21–26 per cent), total (203–267 per cent) and fractional (173–179 per cent) sodium excretion, osmolal clearance (62–111 per cent) and urinary volume (130–161 per cent). These changes were associated with a shift of microspheres from outer to inner cortex, and were completely reversible by eliminating the halothane. In the absence of external influences, halothane produces renal vasodilation and natriuresis. Direct tubular depression cannot be ruled out.