Is there a magical time boundary for diagnosing eyewitness identification accuracy in sequential line‐ups?

We examined whether eyewitness identification latencies for sequential line‐up decisions indicate an optimum time boundary that reliably discriminates accurate from inaccurate decisions. Participants (N = 381) observed a crime simulation and attempted two separate identifications from target‐present or target‐absent sequential line‐ups. As has previously been found with simultaneous line‐ups, the optimum time boundary identified did not reliably discriminate accurate from inaccurate identifications for both line‐up targets. Diagnosticity for choosers was, however, much higher at very high confidence levels than at lower levels. Possible reasons for why one index of signal strength (confidence), but not another (latency), might postdict accuracy within the sequential framework were presented.