Experimental work shows that subretinal fluid is removed both by active transport across the retinal pigment epithelium (RPE) and by passive hydrostatic and oncotic forces that work most effectively when the RPE barrier has been damaged. The retina will stay attached whether or not the RPE is intact--but retinal function requires the RPE barrier and thus active transport is the primary mechanism of subretinal fluid control. RPE fluid transport is normally limited by the retina (which resists water flow from the vitreous) but can be quite powerful when a reservoir of subretinal fluid is present. Clinical serous detachments are unlikely to form solely as a result of small RPE defects or leaks, since the active and passive transport systems for removing subretinal fluid are both so strong. It is suggested that the primary pathology in most serous retinopathy is a diffuse metabolic or vascular abnormality of RPE fluid transport, and that RPE defects or leaks are necessary but only secondary components of the disease. Several hypotheses for removing subretinal fluid therapeutically are considered in terms of their physiology. The subretinal space between the photoreceptors and the retinal pigment epithelium (RPE) is the remnant of the embryonic optic vesicle. In the developed eye the subretinal space is of minimal size, but no tissue junctions form across it and it can re-open under pathological conditions of retinal detachment. In a sense, the title of this paper is misleading since normally there should be no subretinal fluid to control. However, ocular mechanisms are necessary to prevent an accumulation of fluid, and to remove it under conditions of stress or disease.