We read with interest the article by Park et al1 about the use of fibrin sealant in the treatment of anorectal fistulae. We began to treat abscess fistulas with autologous fibrin glue in 1997. All fistulae treated (30 patients) sealed immediately after treatment, but 17 recurred. Recurrent fistulae, treated again by fibrin glue, healed in 4 cases and recurred in 13. The overall success rate, measured by fistula closure, was 56%; this was considered unsatisfactory by the scientific and ethical committee at our institution, and the trial was stopped. When analyzing our results, we came to the conclusion that inadequate removal of the infective source of the fistula could be the cause of our failure. Most fistulae derive from sepses originating in the glands of the anus at the dentate line, and the failure of traditional surgical treatment worldwide is related to residual infection.2 According to the criteria of Parks et al,3 fistulae in our series were classified as intersphincteric in 16 patients (10 with a low and 6 with a high internal orifice), transsphincteric in 10 (6 with a low and 4 with a high internal orifice), and extrasphincteric in 4, all with a high internal orifice. While an effective cleansing of the fistulous tract was attempted in all cases, it is conceivable that in complex fistulae and in those with a high internal orifice, the procedure might have been inadequate.