This study explores how episodic memory (EM) is vulnerable to retroactive interference from working memory (WM). Therefore, we placed a WM task between the encoding and retrieval phases of an EM task. In the EM task, participants learned associations between visual objects and circular locations. An initial recall phase including feedback was then carried out to better consolidate this association. Afterwards, the same objects were used in a WM task, with locations systematically shifted by 120°. We manipulated WM task conditions to investigate retroactive interference mechanisms: In the ‘identical testing’ condition, retrieval formats for both tasks were matched (continuous report of location by changing the object’s circular position). In the ‘deviant testing’ condition, retrieval formats differed between tasks (using spatial placeholders instead of the actual objects for continuous report in the WM task). Additional conditions assessed the effect of shorter versus longer WM storage durations without testing, while the baseline condition involved objects solely used in the EM task. Retroactive interference was measured as reduced precision in EM recall relative to baseline. Results demonstrated significant interference, with the largest drop of precision in the identical testing condition, followed by the deviant testing condition. Conditions without testing exhibited weaker interference effects, irrespective of storage duration. These findings suggest that EM is particularly sensitive to retroactive interference from WM processing when retrieval contexts are similar. Overall, this study challenges the traditional modular perspectives on memory systems and highlights the integrated nature of cognitive mechanisms involved in encoding, maintaining and retrieving memory representations.