Two-dimensional (2D) views dominate the application of linear elastic fracture mechanics to problems in ice-shelf rift propagation, yet from the perspective of fracture mechanics, processes at the rift front are inherently three-dimensional (3D). 2D simplifications are nevertheless desirable for their efficiency and apparent compatibility with shallow-shelf approximation (SSA) ice flow models. Here, the implications of flattening the ice-shelf rift problem are investigated and a theoretical foundation is established to support using a plane stress approximation. In this way, we verify compatibility between 2D simplifications for rifts and for ice-shelf flow, which is a requirement for situations in which SSA-derived stresses are used to study rift propagation. Comparing the plane problem to a 3D counterpart, we show that the 2D results at rift tips are a good estimate for mean rift front stress conditions. Mode I (opening mode) stress intensity factors exhibit a depth-dependence that implies that rifts should be longer at the ice base than at sea level. Other minor 3D effects, which also involve Modes II and III, arise where the rift front intersects the upper and lower ice surfaces.