Abstract Selective attention is central to long-term memory retrieval, yet how attention selects and prioritizes relevant memory contents remains underexplored. Well-controlled experimental designs aimed at isolating and systematically manipulating attentional orienting in long-term memory would greatly advance this cause. The current study addresses this gap. Building on the precue and retrocue designs from working-memory research, we introduce an experimental framework to investigate how selective attention operates within contextual long-term memory. Across three pre-registered experiments ( n  = 235), participants learned associations between scenes and the identities and locations of two objects. In a subsequent Selective Retrieval task, participants viewed the studied scenes and later identified scene-associated objects from an array. Spatial cues directed internal attention to the location of the object to be probed, or neutral cues offered no spatial information. Informative cues yielded significant retrieval benefits and costs relative to neutral cues. Benefits emerged regardless of whether cues appeared before or after scene presentation and were fully expressed even with very short retrieval intervals. The approach offers a flexible method for isolating and manipulating attention functions in long-term memory, enabling future investigation of the psychological and neural mechanisms governing selective memory access.