Objective: HIV-associated neurocognitive disorders (HAND) are associated with deficits in prospective memory (PM). However, most PM research in HIV has used single-event tasks as opposed to habitual PM paradigms, which may be more relevant to clinical populations for whom many health-care behaviors must be performed both frequently and routinely. Method: For the current study, we examined habitual PM and its associations with real-world functioning outcomes in 36 HIV+ individuals with HAND (HAND+), 70 HIV+ individuals without HAND (HAND-), and 115 HIV- individuals. The ongoing task consisted of 24 1-min Stroop trial blocks in which the emotive and cognitive load was manipulated. The habitual PM task required participants to press the spacebar once per block, but only after 20 s had elapsed. Results: A series of MANOVAs covarying for relevant clinicodemographic factors revealed a main effect of study group on habitual PM, such that the HAND+ cohort made significantly more repetition errors than the HIV- and HAND- groups, particularly during early trial blocks. There was no main effect of ongoing task demands, nor was there an interaction between HAND group and task demands. Within the entire HIV+ sample, poorer habitual PM was associated with deficits in learning and dysfunction in real-world outcomes, including medication nonadherence and failures on a naturalistic health-care task. Conclusion: Findings indicate that HAND may be associated with deficient internal source monitoring or temporal discrimination for habitual PM output that may play a critical role in real-world functioning, including HIV disease management.