There is increasing reliance on laboratory testing of new products. The clinical significance of many such studies can be questioned. To function clinically, bond strength must develop more quickly than the shrinkage stress. Oxygen inhibition of extremely thin resin layers prohibits establishment of the bond, perhaps particularly relevant for single-component dentin bonding agents. Use of thicker layers of radiolucent hydrophilic bonding materials can lead to difficulties during subsequent radiographic diagnosis and may contribute to degradation at enamel margins. Shrinkage stress is influenced significantly by many factors. The cavity geometry probably plays the most important role but cannot be controlled to any major extent by the operator. Shrinkage and modulus kinetics are also critical, involving variables such as content and type of monomers, the catalyst system, and the light intensity. In general, a high modulus of elasticity or more rapid polymerization leads to increased shrinkage stress. Combining a single-component adhesive as a dentin primer with a highly radiopaque flowable resin composite as a filled adhesive may help overcome these difficulties.