Purpose To investigate the effects of different transient periods of ischemia and intraocular administration of BDNF on retinal ganglion cell survival. Methods In adult rats, we have applied Fluoro-gold (FG) to both superior colliculi, the main retino-recipient target nuclei in the brain. One week later, when most retinal ganglion cells (RGCs) have incorporated the dye into their cytoplasm, the left retina of several groups of animals underwent transient periods of complete ischemia of 30, 45, 60, 75, 90, 105 or 120 minutes duration by raising intraocular pressure of the eye above systolic levels. The rats were perfused through the heart 5, 7, 15 or 30 days later, their retinas prepared as flattened whole-mounts and examined under fluorescence microscopy for FG-labelled RGCs. Mean densities of surviving (FG-labelled) RGCs were estimated by counting these cells in twelve standard areas of the retina according to previously described methods (Villegas-Perez et al., J. Neurobiol. 1993, 24:23). In additional groups of animals, right after a transient ischemic period of 90 minutes, the eye was punctured or injected with 5 μl of saline containing 0,1% BSA alone or with 5 μg of BNDF, and their retinas processed as above 7 days later.Results In the groups of rats analyzed at 5, 7, 15 and 30 days we found that: i) Periods of ischemia below 45 minutes did not cause significant decreases of RGC densities; ii) After periods of ischemia of 60 minutes, RGC survival was 67%, 63%, 51% and 45% respectively. After periods of ischemia of 90 minutes, RGC survival was 52%, 43%, 16% and 4% respectively. iii) After periods of ischemia of 105 and 120 minutes, only a small population of RGCs survived 15 and 30 days after ischaemia. iv) The groups of retinas treated with BDNF, vehicle or sham injection showed moderate increases in FG-labelled RGC densities when compared to untreated retinas.Conclusions In the adult rat retina: i) Transient periods of ischemia of 30 and 45 minutes do not induce RGC death; ii) Longer periods of ischemia induce the death of a proportion of RGCs that increases with the duration of the period of ischemia; iii) Periods of transient ischemia above 90 minutes induce the death of approximately 95% of the RGC population; iv) RGC death can be observed as soon as 5 days after transient ischemia, but continues during the period of this study (up to 30 days), and; v) Intraocular injection neuroprotects a small population of RGCs.