Microleakage of oral microorganisms, which can occur due to the lack of sealing ability of permanent restorative materials, may cause failure of root canal treatments. Although a great deal of research has been done on sealing enamel and coronal dentin with resins, little research has been done on the adhesion of resins to the walls of pulp chambers. The purpose of this study was to evaluate regional bond strengths of two adhesive systems to the walls of pulp chambers. A section was made horizontally through the middle of the pulp chamber of extracted human third molars to divide the chamber into upper and lower halves. The pulp tissue was removed and the tooth segments were then divided into treatment subgroups. The pulp chambers were bonded with C&B Metabond (Parkell) or One-Step (Bisco), with or without 5% NaOCl pretreatment. The microtensile bond strengths of these resins to four different pulp chamber regions (bottom, wall, roof, and pulp horn areas) were then measured using an Instron machine. The data were expressed in MPa and were analyzed by a three-way ANOVA. Statistically significant differences were found among the test groups (p < 0.001). One-Step produced higher bond strengths to all pulp chamber regions except the floor, compared with C&B Metabond. The results indicated that high bond strengths can be achieved between adhesive resins and the various regions of the pulp chamber. This should permit the use of a thick layer of unfilled resin along the floor of the pulp chamber and over the canal orifices as a secondary protective seal after finishing root canal therapy. Copyright © 2001 by The American Association of Endodontists.