Witnesses to crime often experience stress during the witnessed event. However, most laboratory studies examining eyewitness memory do not include a stressful encoding event. Participants (N = 129) completed an experimental stress induction procedure--a modified version of the Trier Social Stress Test. We designed three conditions to manipulate the amount of stress experienced and included three types of measures to assess the effectiveness of the manipulation: cortisol levels (hormonal), blood pressure and heart rate (autonomic), and self-report (subjective). Participants watched a video that had a surprise viewing of a staged theft and completed two lineup identification tasks. We observed no effects of stress on the accuracy or willingness to choose from a lineup. Importantly, there was variability in the correspondence between measured indicators of stress, which should be considered in future designs.