The objective of this study was to optimize recovery and cryopreservation of epididymal sperm from plains bison, as a model for wood bison. In Phase 1, cauda epididymides were recovered from bison (n = 14) immediately after slaughter, minced and incubated in Sp-TALPH buffer for 3 h at 36 degrees C. The resulting sperm suspensions were cryopreserved in Triladyl, using a protocol for bovine semen. In Phase 2, epididymal sperm were cryopreserved in either Triladyl or Andromed. The mean (+/-SD) estimated number of sperm recovered was 468 +/- 207 x 10(6). There was an increase (p < 0.05) in the proportion of sperm with normal morphology between initial recovery and after extension (52.4 +/- 4.6 vs 69.7 +/- 2.4%), with a concurrent decrease (p < 0.05) in the proportion of sperm with distal droplets. Median values for progressively motile sperm in post-thaw samples (60%) were lower (p < 0.05) than that after extension or after chilling (70% for both). The mean percentages of viable sperm and of sperm with an intact acrosome were lower (p < 0.05) for frozen-thawed samples (38.7 +/- 2.8 and 85.2 +/- 1.1) compared with extended (66.2 +/- 2.2 and 92.4 +/- 0.9) or chilled (63.7 +/- 2.5 and 90.0 +/- 1.0) samples. Rates of cleavage, morulae and blastocyst production were not significantly different for chilled (70.9, 38.7 and 8.0%) vs post-thaw sperm (73.0, 46.0 and 6.3%). There was no significant difference between extenders for most sperm characteristics. In conclusion, we developed a functional protocol for the recovery and cryopreservation of epididymal sperm from plains bison, which may have implications for the genetic preservation of wood bison.