The optimization of cryopreservation extenders is a fundamental issue for adequately performing germplasm banking on wild species. We have tested two glycerol concentrations (4 and 8%), and three extender osmolalities (320, 380 and 430 mOsm/kg; before adding cryoprotectants), for cryopreservation of epididymal and ejaculated sperm samples from Iberian red deer. All the extenders were based on Tes-Tris and fructose (for osmolality adjustment), and complemented with 20% egg yolk. Epididymal and ejaculated sperm samples were obtained from the cauda epididymis (post-mortem) and using electroejaculation, respectively. Samples were diluted 1:1 with each extender and equilibrated for 2 h at 5 degrees C. Then, they were diluted down to 100x10(6) sperm/mL and frozen at -20 degrees C/min. Post-thawed samples were assessed for motility (CASA), HOS test, proportion of swollen (osmotically challenged) cells in the untreated sample, viability and acrosomal status. For epididymal samples, 8% glycerol rendered a slightly higher proportion of intact acrosomes on viable spermatozoa than 4%; regarding extender osmolality, 380 and 430 mOsm/kg rendered higher motility results, and the 430 mOsm/kg yielded the lowest proportion of swollen spermatozoa. For ejaculated samples, 4% glycerol yielded more viable spermatozoa than 8%; for extender osmolality, 320 mOsm/kg rendered the highest percentages of progressively motile and viable spermatozoa, although 380 mOsm/kg extender was not significantly different. These results show that sample source influences extender suitability, and that extenders should be isoosmotic or rather slightly hyperosmotic. Future studies should test multiple glycerol concentrations and extender osmolalities in order to adjust them to these kinds of sample.