In three experiments we examined individual differences in working memory (WM) and their relationship with filtering—the selective encoding and maintenance of relevant information in the presence of irrelevant information. While some accounts argue that filtering is an important element of individual differences in WM (McNab & Klingberg, 2008; Robison & Unsworth, 2017b; Unsworth & Robison, 2016; Vogel, McCollough, & Machizawa, 2005), recent investigations have challenged this view (Mall, Morey, Wolff, & Lehnert, 2014; Shipstead, Lindsey, Marshall, & Engle, 2014). In all three experiments, we measured WM span with three complex span tasks and then had participants complete a visual WM task with a filtering component. In Experiment 1, participants were instructed to remember the orientation of relevant items (red rectangles) and ignore irrelevant items (blue rectangles). In Experiment 2, the color of relevant items changed randomly on a trial-by-trial basis. In Experiment 3, we presented a constant number of items. On half of the trials, participants were told which color item would be tested before each trial. On the other half of the trials, participants received no such cue. In situations where filtering was especially required, WM span accounted for a significant portion of variance in filtering trials beyond shared variance between filtering and nonfiltering trials. We argue that filtering is one of several control processes that gives rise to individual differences in WM, but that the relationship is constrained by the degree to which filtering is demanded by the task.