Moisture-assisted crack growth in polymer adhesive/glass interfaces was measured as a function of the applied energy release rate, G, using a four-point flexure test coupled with an inverted microscope. The specimens consisted of two glass plates bonded together with an epoxy or an epoxy-acrylate adhesive. It was found that cracks formed and grew on both interfaces if the glass surfaces were both smooth; however, roughening the surface of one of the glass plates increased the fracture resistance of the interface sufficiently so that crack growth occurred only on the remaining ＇＇smooth＇＇ interface (top or bottom). Finite element analysis was used to determine the G and psi (phase angle) appropriate for the different crack geometries. It was found experimentally that crack growth rates for all crack geometries depended on the applied G via a power law relationship and that for a given applied G, crack growth rates were sensitive the crack geometry. The results indicate that the primary driving force for moisture-assisted crack growth at a polymer/glass interface is the applied G at the crack tip and that the effect of the phase angle for the different crack geometries (13 degrees to 54 degrees) is negligible.