We have recently presented a new technique for occlusion of long standing fistulae by excochleation and plugging with a fibrin clot (Lancet, 1979). The intention with the technique was to keep the fistulous tract dry and occluded by a species-specific substance until healing had occurred. To study the importance of the fibrin clot in the present context an experimental study in rats has been performed. Perforated Teflon cylinders were implanted subcutaneously, removed after 1, 2 and 4 weeks and subjected to macro-and microscopic examination. Comparison was made between empty cylinders and cylinders filled with a fibrin clot. The examination after one week revealed the initially fibrin clot filled cylinders still to be filled by the clot. Ingrowth of granulation tissue through the cylinder perforations had started. The initially empty cylinders were still empty. At the 2-week examination the initially fibrin clot filled cylinders were at the peripheral half filled by ingrowing granulation tissue, the central half was filled by the implanted fibrin clot. The initially empty cylinders were filled to a lesser degree by ingrowing granulation tissue and a loose exdogenous clot centrally. The 4-week examination revealed both the initially empty and the initially fibrin clot filled cylinders to be completely filled up by granulation tissue of varying organisation. During the time it takes for the granulation tissue to grow into the cylinder the clot remains intact centrally and together they keep the cylinder filled up. It furthermore appears as if the fibrin clot induces a more rapid ingrowth into the cylinder. Both observations confirms the importance of the clot for the fistula occlusion.