Abstract
Enamel-surrounded cavities prepared in extracted human teeth were treated with an enamel-bonding agent, one of two commercial dentin adhesives (Scotchbond 2 and Tenure), or one of two recent modifications of the Gluma system. In all cases the enamel margins were subjected to some form of acid etching. The cavities were filled with one of two different light-cured posterior resins, and the wall-to-wall polymerization contraction was assessed in the enamel surface level and, after grinding, measured and presented as linear shrinkage in the dentin surface level. The linear shrinkage varied from 0.66% in the control group, in which only composite resin was applied, to 0.28% in a group in which the cavities were pretreated with a dentin adhesive. No difference could be demonstrated between the dentin adhesives or between the dentin adhesives and the enamel-bonding agent.