The orbicularis oculi branch of the facial nerve was transected in two groups of 15 cats. Following section of the nerve, a specimen was removed for histological studies; then, using microscopic techniques the epineurium was sutured (nylon 10:0) in the first group, and the perineurium in the second. They were then followed for 100 days. Results rely on clinical (complete closure of the eye), electrophysiological (summation potentials), and histological (fiber count) evaluations. Judging from the relative density and diameters of the fibers, epineural sutures result in better regeneration than perineural sutures. With the tremendous importance of the absence of tension on the suture line in mind, it seems that the technique of choice for monofascicular nerves, as in the pyramidal and tympanic segments of the facial nerve, is epineural suture. Despite the slight advantage of epineural suture over perineural suture as evidenced by histological evaluation, there was no clinical difference. The distal fascicular distribution of facial nerve fibers and their spatial distribution make perineural (fascicular) sutures the method of choice when anastomosis becomes necessary in the vertical portion and in the portion distal to the stylomastoid foramen.