Results are presented from an experimental investigation of the effects of stacking sequence on the delamination toughness and delamination growth behavior exhibited by laminated composite end-notched flexure (ENF) test specimens. Delamination growth is studied at interfaces defined by 0°/0°, +15/−15°, +15/+15°, +30/−30°, and +30/+30° plies in a typical graphite/epoxy. A previously developed “ENF test design procedure” is used to aid in the determination of specimen stacking sequences and test geometries, such that delamination growth occurs at the interface of interest, yet the contributions to the energy release rate from residual thermal stresses, geometric nonlinearities, local Mode II concentrations at the specimen's free edges, and local Mode III effects are minimized. Critical strain energy release rates are obtained from the load-deflection test data by both a classical plate theory-based data reduction technique and by a compliance calibration procedure and the results compared. It is shown that compliance calibration is the only completely reliable means of reducing the data from either unidirectional or multidirectional ENF tests. It is also shown that the proposed ENF test design procedure reduces or eliminates the propensity for subcritical delamination growth at the free edges of an ENF specimen. For specimens that are not precracked, there is no significant effect of interfacial ply orientation on GIIc for the material and relative angles studied.