Abstract. The time-dependent mechanisms of the glass transition can be exploited to program large recoverable shape changes in amorphous polymers. These shape memory polymers have enormous potential applications as morphing structures and deployable structures for aerospace and biomedical applications. The shape memory response in polymers is robust and repeatable, but the performance of current shape memory polymer devices are limited by the material's slow response time and small activation force. The shape memory performance is controlled by a combination of many factors, including thermomechanical properties and programming conditions. We developed a constitutive model to predict the shape memory behavior of amorphous networks that featured the time-dependent viscoelastic and structural relaxation mechanisms of the glass transition. This presentation describes the model formulation, experimental methods developed to measure the stress and structural relaxation behavior for parameter determination, and application of the model to predict the recovery performance under a variety of mechanical constraints.