Upper extremity prostheses are designed to meet a spectrum of user needs in the areas of function and cosmesis. Nevertheless, high rates of prosthesis abandonment suggest that we do not sufficiently understand consumer satisfaction, design priorities and the critical factors in device acceptance/rejection. This thesis establishes a user-based design framework to more closely connect the consumer, clinician and designer. The framework is exemplified by the profiling of individuals with limb absence and the subsequent evaluation of electroactive polymers as a potential solution to identified consumer priorities. To this purpose, a comprehensive, international survey of 262 participants was conducted. Multi-variate models (i.e. logistic regression, decision trees, radial basis function networks) suggest four primary factors in prosthesis rejection:​ fitting time-frame, involvement in prosthesis selection, age, and level of limb absence. Model specificity (prediction of wearers) and sensitivity (prediction of rejecters) of up to 93% and 80% were obtained, respectively. Design priorities focus on comfort, particularly improvements in weight, heat dissipation, fit, and harnessing options. Life-like appearance/function and sensory feedback are also of interest, particularly for children and wearers of passive or electric hands. To meet with consumer design priorities, the potential of light-weight dielectric elastomers as muscle-like actuators was empirically characterized with respect to performance, energy requirements, durability, and controllability. Challenges in these areas (i.e. low strain at high frequencies, insufficient power density, limited fatigue life, sensitivity to contaminants, and presence of hysteresis/creep), suggest the need for fundamental material developments before implementation in upper limb prosthetics is feasible. To fulfill consumer desire for sensory feedback, the performance of ionic polymer metal composites as light-weight, flexible, compact bend sensors was also explored. With linear responses within the operating range typical of hand prostheses, bending angles and rates were accurately predicted with errors of 4.4 +/- 2.5% and 4.8 +/- 3.5% respectively, comparable to traditional sensors. Future work in upper limb prosthetics calls for the union of clinical knowledge, engineering expertise and consumer experience in order to provide the best possible resources to individuals with upper limb absence. The research presented in this thesis provides a framework that facilitates such inter-professional, consumer-centric collaboration.