Past research has consistently shown that episodic memory (EM) declines with adult age and, according to the associative-deficit hypothesis, the locus of this decline is binding difficulties. We investigated the importance of establishing and maintaining bindings in working memory (WM) for age differences in associative EM. In Experiment 1 we adapted the presentation rate of word pairs for each participant to achieve 67% correct responses during a WM test of bindings in young and older adults. EM for the pairs was tested thereafter in the same way as WM. Equating WM for bindings between young and older adults reduced, but did not fully eliminate, the associative EM deficit in the older adults. In Experiment 2 we varied the set size of word pairs in a WM test, retaining the mean presentation rates for each age group from Experiment 1. If a WM deficit at encoding causes the EM deficit in older adults, both WM and EM performance should decrease with increasing set size. Against this prediction, increasing set size did not affect EM. We conclude that reduced WM capacity does not cause the EM deficit of older adults. Rather, both WM and EM deficits are reflections of a common cause, which can be compensated for by longer encoding time. (PsycINFO Database Record (c) 2018 APA, all rights reserved).