Recent research by Barrouillet et al. (2004, 2007) have proposed a shared-resource model of working memory (WM) detailing a detrimental time-based bottleneck effect of memory retrievals for simultaneous processing and storage tasks. This Time-Based Resource-Sharing Model directly contradicts previous research by Duff and Logie (1999, 2001) into the separation of processing and storage components of WM, and as a result has wider implications concerning the structure of WM. The experiment in this paper was designed in an attempt to test the claims made by Barrouillet et al. (2007) concerning the retrievals made during a processing task and their effect on concurrent memory span. An experimental paradigm was developed that is directly comparable to previous research in this field, yet allowed manipulation of the retrieval difficulty of the processing task stimuli. The results of this experiment followed a similar pattern to Duff and Logie's (1999, 2001) research and found no significant effect of processing task retrieval difficulty on concurrent memory span as well as no detectable slowdown of reaction time during the processing task. The implications of these findings on Barrouillet et al.'s (2004, 2007) TBRS model are explored, along with a discussion of possible future research in this area.