Previous studies have demonstrated that instructional manipulation of a participant witness’s response criterion on a description task can lead to verbal overshadowing in performance on a subsequent lineup identification task. The current set of experiments attempts to replicate this instructional bias effect in verbal overshadowing and extend the paradigm to include variations in lineup presentation format (Experiment 1) and repeated descriptions prior to identification (Experiment 2). The instructional bias effect is found to persist despite the “sequential” presentation of lineup members and across repeated recall of a target face 1 week later. Furthermore, both experiments demonstrated that incorrect details generated by participants were predictive of subsequent inaccuracy on the identification task. Both theoretical and applied aspects of the instructional bias effect in verbal overshadowing are discussed.