Objectives/Hypothesis (1) To measure attentional selection based on the three-network model of attention (alerting, orienting, and executive control) in older adults with and without hearing loss, and (2) in adults with hearing loss (a) awaiting a CI and (b) currently using a CI. The principal hypothesis was that hearing rehabilitation with CIs would release higher-order explicit cognitive processes, resulting in attentional selection among CI users similar to controls. Study Type Clinical. Design Cross-sectional, between-groups. Methods Attentional selection was measured using the nonauditory, visual Attentional Network Test (ANT) in (1) normal hearing older adults, (2) age-matched postlingually deaf (2) CI candidates, and (3) CI users. Results Eleven normal hearing controls, 12 CI candidates, and 16 CI users comprised our sample. ANT performance differed significantly between normal hearing controls and participants with hearing loss. Both CI candidates and CI users made less use of temporal alerting cues compared to normal hearing controls. In contrast, CI candidates and CI users utilized spatial orienting to a greater extent than normal hearing controls. No significant differences were noted among controls, CI candidates, or CI users in executive control. Conclusions Attentional selection and efficiencies differ between older adults with and without hearing loss. Participants with hearing loss utilized spatial cues preferentially and temporal cues minimally. No significant differences in attentional selection were observed between CI candidates and CI users. Level of Evidence 3.