This study examined whether a sign recall task on a driving simulator, self-report of driving ability, or age predicted differences in performance between drivers with mild cognitive impairment (MCI) and control participants. For the dependent measure, gathered using a driving simulator, working memory was subjected to interference at varying levels of driving task demands. Reliable between-groups differences in sign recall accuracy were demonstrated; recall declined under higher task demands. Recall scores, self-reported frequency of avoiding driving, and driver age did not predict MCI; only self-reported decline in global driving ability was significant. Findings support the use of driving simulators in practice and suggest that screening for age-related cognitive impairment should incorporate self-reported changes in driving proficiency for early identification of drivers who merit medical review. The results, although exploratory, have implications for practitioners.