If several items are associated with a common cue, the cued recall of an item is often supposed to decrease as a function of the increase in strength of its competitors' associations with the cue. Evidence for such a list-strength effect has been found in prior research, but this effect could have been caused both by the strength manipulations and by retrieval-based suppression, because the strengthening and the output order ofthe items were confounded. The experiment reported here employed categorizable item lists; some categories in each list contained strong items only,some contained weak items only, and some contained both strong and weak items. Strengthening was accomplished by varying the exposure time of the items. The testing sequence of the items from each category was controlled by the use of category-plus-first-letter cues. When the typical confounding of strengthening and output order was mimicked,list-strength effects were found, which is consistent with prior research. However, when this confounding was eliminated, the list-strength effects disappeared: The recall of neither strong nor weak items varied with the strengths of the other category exemplars. This pattern of results indicates that the list-strength effect is not the result of strength-dependent competition, but is caused by output-order biases and a process of suppression.The cued recall ofan item decreases as the number of items associated with the same cue increases. Corresponding evidence has been provided by a number of studies in quite different experimental paradigms (see Watkins, 1978, for a review). This finding is generally seen as support for the assumption that memories associated with a common cue compete for access to conscious recall when that cue is presented.Often, this competition is assumed to be strength dependent; that is, the cued recall of an item is supposed to decrease as a function of the increase in strength of its competitors' associations with the cue. Evidence for this strength dependence arises particularly from studies on retroactive and proactive interference (Barnes & Underwood, 1959;Briggs, 1957), although the results from studies on part-set cuing have been interpreted similarly (Roediger, 1974;Rundus, 1973). This evidence for strength dependence appears to be so compelling that the concept has been incorporated as a fundamental property into many memory models (Mensink & Raaijmakers, 1988;Raaijmakers & Shiffrin, 1981;Rundus, 1973).The concept of strength dependence, however, is far less well empirically established than it would seem. Recently, Anderson, Bjork, and Bjork (1994) emphasized this point by arguing that previous interference studies, which seemingly provide evidence for strength dependence, confounded the manipulation of the strengths of Thanks are extended to M. Anderson, A. Glenberg, H. L. Roediger III. and one anonymous referee for their comments on an earlier draft of this manuscript. The author's address is Institut fur Psychologie, Universitat Regensburg, 93040 Regensburg, Germany (e...