Prior research focused on the relationship between cognitive offloading and working memory ability in the prospective and retrospective memory domains have produced conflicting results. Specifically, past work in the prospective memory domain has found that individuals with lower working memory capacity (WMC) choose to offload more often and benefit more from offloading than those with higher WMC (Ball, Peper, et al., 2022) while work in the retrospective memory domain has not found a relationship between WMC and the use of or benefit from offloading (Morrison & Richmond, 2020). However, task design across studies differed in several other respects aside from memory domain, making it difficult to discern whether different mechanisms underlie cognitive offloading across domains. The current study aimed to address these discrepancies by introducing similar procedures across offloading tasks. Results revealed that when offloading was required or permitted, participants with varying levels of WMC generally performed more similarly to one another than when the task had to be completed using internal memory alone. In addition, participants with lower WMC generally benefitted more from offloading, particularly under high memory load, compared to those with higher WMC when offloading was required and when participants had free choice about whether and when to engage in offloading. However, neither metacognitive underconfidence in internal memory capability nor lower WMC estimates were associated with increased offloading frequency in either memory domain when participants were permitted to offload. Practical and theoretical implications of these findings are discussed.