The discrepancy in the choroidal circulation between anatomy and function has remained unsolved for several decades. Postmortem cast studies revealed extensive anastomotic channels, but angiographic studies indicated end-arterial circulation. We carried out experimental fat embolism in cats and electric circuit simulation. Perfusion defects were observed in two categories. In the scatter perfusion defects suggesting an embolism at the terminal arterioles, fluorescein dye filled the non-perfused lobule slowly from the adjacent perfused lobule. In the segmental perfusion defects suggesting occlusion of the posterior ciliary arteries, the hypofluorescent segment became perfused by spontaneous resolution of the embolism without subsequent smaller infarction. The angiographic findings could be simulated with an electric circuit. Although electric currents flowed to the disconnected lobule, the level was very low compared with that of the connected ones. The choroid appeared to be composed of multiple sectors with no anastomosis to other sectors, but to have its own anastomotic arterioles in each sector. Blood flows through the continuous choriocapillaris bed in an end-arterial nature functionally to follow a pressure gradient due to the drainage through the collector venule.The choroid is a very special tissue, as it has the most abundant blood flow per weight in the human body and a highly complicated three-dimensional vascular structure. Choroidal circulation has been suggested to be predisposed to many diseases, including age-related macular degeneration 1, 2 , polypoidal choroidal vasculopathy A majority of our knowledge about choroidal circulation has been obtained from classical studies using autopsy and angiography. Postmortem corrosion cast studies revealed that choroidal circulation has anastomotic channels at various levels of capillaries, arterioles, and venules 6,7 . These characteristics are shared with various mammals including monkeys 6 , horses 8 , and cats 9 . In contrast, extensive angiographic studies conducted in vivo by Hayreh 10-12 indicated that the choroid has an end-arterial system. Recent advent of choroidal imaging using new technologies, such as Doppler optical coherence tomography and continuous laser-targeted angiography, affirmed the previous observations, leaving the disparity unresolved [13][14][15] . Although an autonomic nerve supply of the choroid in vivo was suggested as explanation, the discrepancy between the anatomy and the function has remained unsolved for several decades 10,12 . Flower and colleagues 16,17 suggested that the blood follows a pressure gradient in vivo based on the observations that choriocapillaris blood flow was inhomogeneous and variable upon changes in blood pressure and intraocular pressure. Although the vascular channels are interconnected in the choroid, the blood will follow pressure gradient and will be observed as an end-arterial system in angiography. A "functional end-arterial system" represents the above theory 15 . However, the theory has not been...