Although general anesthesia frequently is used for eye surgery or used in experimental studies of circulation in the eye, few data are available describing its effects on ocular blood flow. The blood supply to the retina in humans and other mammals is derived from a dual circulation; the retinal vessels supply the inner neural layers while the choroidal vessels supply the outer retina. Both circulations are required for normal retinal function. Using radioactively labeled 15-mu-m microspheres containing Ce-141, Sn-113, or Nb-92, blood flow was measured in the retina and choroid in cats (whose ocular circulation is similar to that of humans) and cerebral cortex during halothane anesthesia. In ten adult cats, retinal blood flow was 37 +/- 3, 54 +/- 6, and 59 +/- 4 ml . 100 g-1 . min-1 (mean +/- SEM) at 0.5, 1.0, and 1.5 MAC halothane, respectively, and corresponding values for cerebral cortical blood flow were 60 +/- 5, 69 +/- 6, and 98 +/- 14 ml . 100 g-1 . min-1 (mean +/- SEM), respectively. For both retinal and cerebral blood flows, values obtained at 1.0 and 1.5 MAC were significantly greater than those at 0.5 MAC (P < 0.0167). In contrast to the effects on retinal blood flow, choroidal blood flow was significantly decreased during halothane anesthesia. Choroidal blood flow was 1,801 +/- 222, 1,309 +/- 167, and 1,091 +/- 126 ml . 100 g-1 . min-1 (mean +/- SEM) at 0.5, 1.0, and 1.5 MAC, respectively. Values obtained at 1.0 and 1.5 MAC differed significantly from those at 0.5 MAC (P < 0.0167). Increasing concentrations of halothane resulted in significant decreases in ocular perfusion pressure (mean arterial pressure minus intraocular pressure) from 118 +/- 8 at 0.5 MAC to 99 +/- 7 and 78 +/- 7 mmHg (mean +/- SEM) at 1.0 and 1.5 MAC, respectively (P < 0.0167). Decreases in choroidal blood flow and ocular perfusion pressure were parallel. It is concluded that halothane anesthesia produces a significant dose-dependent increase in retinal, and a decrease in choroidal blood flow in cats.