Although there is some evidence that setting performance criteria may improve oral reading fluency interventions, little is known about the generalized effects of these criteria. The present study trained two third‐grade students to three different fluency levels on instructional passages and assessed generalized performance in corresponding high word‐overlap passages within an alternating treatments design. Results indicated no discernable differences across experimental conditions. Follow‐up analyses revealed that generalized gains per trial were actually larger in lower criterion conditions, suggesting that performance criteria may not be as helpful as previously thought. The results are discussed in terms of the need to empirically investigate the number of instructional trials necessary to maximize instructional efficiency. Copyright © 2016 John Wiley & Sons, Ltd.