Brown bear is a highly endangered species in Spain and could benefit from biological resource banking. Currently, we are studying several reproductive aspects in order to aquire the knowledge for establishment of a germplasm bank for this species. One of our objectives is to develop adequate protocols for the evaluation of bear sperm before and after cryopreservation. We have used the fluorescent probe JC-1 protocol, which differentially stains mitochondria, according to its activity (Garner DL et al. 1997 Biol. Reprod. 57, 1401–1406). Here we describe one problem that arose using this staining for evaluation of extended bear semen. We electroejaculated 13 adult brown bears (Ursus arctos) (206–311 kg) housed in a half-freedom regime in the Cabarceno Park (Cantabria, Spain). Anesthesia was performed with tiletamine + zolazepan (Zoletil 100®, 7 mg/kg; Virbac, Spain), and ketamine (Imalgene 1000®, 2 mg/kg; Mericl, Sain). We used an electroejaculator (PT Electronics®; Boring, OR, USA) with a 3-electrode transrectal probe (26 mm in diameter, 320 mm long). Ejaculation occurred at 10 V/250 mA. Samples were extended (prepared in our laboratory, Anel L et al. 2003 Theriogenology 60, 1293–1308; M3 modified) and cooled to 5°C for 70 min (pre-freezing protocol). We analyzed individual (MI) and progressive (MP) motility by means of an automated motility analyzer (Hamilton Thorne Biosciences, Inc., Beverly, MA, USA), using a phase contrast microscope (Nikon, ×10). Mitochondrial status was analyzed after diluting the sample 1:100 with buffered medium (20 mM HEPES, 153 mM NaCl, 2.5 mM KOH, 10 mM glucose; Sigma, Madrid, Spain) and adding JC-1 (6.8 μM final; Molecular Probes, The Netherlands). After 30 min at 37°C, 100 cells were counted with an epifluorescence microscope (Nikon, ×400), determining the percentage of sperm with orange-stained (active) mitochondria. We analyzed a total of 55 samples in three different models: fresh, pre-freezing, and thawed. We divided the samples into successful JC-1 staining (valids: V) or failed JC-1 staining (not valid: NV) (depending on the aspect of the stained cells). In not-valid samples we observed a greenish background, with almost no fluorescent spermatozoa. These observations were consistent in a given sample, giving the same V or NV result when we repeated the staining. In fresh and thawed groups there were no NV samples, but in the pre-freezing group there were 40 NV samples (73%). We calculated Pearson correlations (SAS; SAS Institute, Inc., Cary, NC, USA) between percent JC-1 orange population and MI and MP in fresh (r = 0.40 and 0.33; P < 0.001), thawed (r = 0.61 and 0.43; P < 0.001) and pre-freezing samples (r = −0.11 and −0.24; P > 0.05), all respectively. When pre-freezing samples were split between V and NV, the former had good correlations (r = 0.74 and 0.49; P < 0.05), and NV still did not (r = −0.17, −0.27; P > 0.05). We conclude that JC-1 staining is not reliable for the pre-freezing analysis of bear sperm, at least under the conditions described here. This could be due to the interaction of the extender or the refrigeration treatment with the sperm. However, this problem did not occur in the analysis of fresh and thawed samples. Nevertheless, it may be advisable to test other mitochondrial probes for analyzing this kind of samples.