The murine retinal explant is a popular model for the study of neuronal degeneration and repair (1). Therapeutic efficacy is usually quantified by cell counts/viability markers, which provide limited information regarding mechanisms leading to cell death. Predicated on the evidence that dendropathy precedes cell loss in most neuronal degeneration (2, 3), we postulated that in the retinal explant dendrite analysis may provide a more sensitive measure of neuronal health than cell counts alone. Explants from adult C57/Bl6 mice were cultured as wholemounts at 37°C, 5% CO2 for up to 14 d. Frozen sagittal sections were nuclear stained and immunohistochemically stained for the apoptotic marker active caspase-3. Retinal ganglion cells (RGCs) were labelled diolistically (DiI/DiO) for the quantification of dendritic arbours by Sholl analysis (4). The effect of brain-derived neurotrophic factor (BDNF) was investigated by incubation with BDNF (100 ng/mL) for 3 d initiated at 0 d (3 d total) or at 3 d (6 d total). Values are means ± S.E.M.. Nuclei counts, BDNF Sholl area under the curves (AUCs) and non-BDNF branching indexes were compared by ANOVA with TUKEY post-hoc, active caspase-3 staining and non-BDNF Sholl AUCs were compared by Mann-Whitney with Bonferroni correction, and BDNF branching indexes were compared by independent samples T-test. The ganglion cell layer (GCL) nuclei count significantly decreased by 14 d (65±6 vs.104±5 cells/mm at 0 d, p<0.005, n=18 retinas) with a corresponding 45.6% increase in active caspase-3 in the GCL (2.69±1.24 vs. 1.85±0.52 +ve cells/mm at 0 d, p>0.05, n=16 retinas). Significant dendropathy was observed at 1 d, quantified as a 36.6% reduction (1136±94 vs.1737±116 at 0 d, p<0.001) in Sholl AUC and a 43.0% reduction (645±66 vs. 1133±95 at 0 d, p<0.05) in branching index (n=178 cells). BDNF treatment resulted in a 61.5% increase (1365±180 vs. 845±134 for control, p<0.05) in Sholl AUC and an 80.3% increase (909±130 vs. 504±83 for control, p<0.05) in branching index relative to controls (n=36 cells). Delayed BDNF treatment resulted in a 136% increase in Sholl AUC (1166±147 vs. 495±74 for control, p<0.001) and a 107% increase (610±86 vs. 295±43 for control, p<0.005) in branching index relative to controls (n=36 cells). The delayed BDNF Sholl AUC increased relative to 3 d Sholl AUC (56.6%,1166±147 vs. 745±82 for 3 d, p<0.005). We report that in adult mouse retinal explants, RGC dendropathy precedes cell loss by at least 7 d. We demonstrate that dendrite analysis provides a sensitive method for monitoring neuronal health and a sensitive readout for the effects of therapeutic interventions. This assay facilitated the identification of a treatment window when neuronal cell structure and function may be recovered.