Although the retina has one of the highest metabolic rates among tissues, certain teleost fishes lack any vascular supply to this organ which, in combination with the overall thickness of the organ, results in extremely long diffusion distances. As the only way to compensate for these obstacles, oxygen partial pressure (PO2) in the eyes of such fish is elevated far above atmospheric values. Although not supported by any direct evidence, the enhancement of PO2 is considered to be related to the Root effect, the release upon acidification of Hb-bound O2 into physical dissolution, possibly supported by counter-current multiplication similar to the loop of Henle. The present study evaluates the magnitude of intraocular PO2 enhancement under tightly controlled physiological conditions, to directly confirm the involvement of the Root effect on intraocular PO2 in the retina of rainbow trout Oncorhynchus mykiss. Intraocular PO2 was determined with special polarographic microelectrodes inserted into the eye. PO2 profiles established in vivo by driving electrodes through the entire retina yielded average PO2 values between 10 mmHg (1.3 kPa) at the inner retinal surface and 382 mmHg (50.9 kPa) close to the outer retinal limit (Bruch's membrane). According to estimates on the basis of the diffusion distances determined from sections of the retina (approximately 436 microm at the site of PO2 measurement) and literature data on specific oxygen consumption, the in vivo determined values would be sufficient to cover the oxygen demand of the retina with some safety margin. For a clear and direct in-tissue-test as to the involvement of the Root effect, an isolated in vitro eye preparation was established in order to avoid the problem of indirect blood supply to the eye from the dorsal aorta only via the pseudobranch, a hemibranch thought to modulate blood composition before entry of the eye. Any humoral effects (e.g. catecholamines) were eliminated by perfusing isolated eyes successively with standardized red blood cell (RBC) suspensions in Ringer, using trout (with Root) and human (lacking any Root effect) RBC suspension. To optimize perfusate conditions for maximal Root effect, the Root effect of trout RBCs was determined in vitro via graded acidification of individual samples equilibrated with standardized gas mixtures. During perfusion with trout RBC, PO2 at the outer retinal limit was 99 mmHg (13.2 kPa), but fell by a factor of 3.3 upon perfusion with human RBC in spite of higher total oxygen content (TO2 2.8 for trout vs 3.9 mmol l-1 for human RBC). Upon reperfusion with trout RBC, PO2 was restored immediately to the original value. This regularly observed pattern indicated a highly significant difference (P=0.003) between perfusion with trout (with Root effect; high retinal PO2) and perfusion with human (no Root effect; low retinal PO2) RBC suspension, thus clearly demonstrating that the Root effect is directly involved and a crucial prerequisite for the enhancement of PO2 in the retina of the teleost eye.