This dissertation reports 4 Experiments that are concerned with the evaluation of eyewitness identification testimony. In Experiment 1, three target description groups were assessed in order to test different theoretical accounts regarding the relationship between identification performance and target description (verbal overshadowing): non-describers, describers, and rereaders (describers with rereading of the description before the identification task). Rereaders less frequently chose somebody from the lineup than the other two groups, lending support to the decision criterion shift approach (Clare & Lewandowsky, 2004). In Experiment 2, post-decision confidence, decision time, and self-reported decision processes were used as postdictors of identification accuracy. Using a decision rule including highly confident and fast choosers led to more correct classifications than either variable alone. Unexpectedly, self-reported decision processes were neither associated with identification accuracy for choosers nor for nonchoosers. In Experiment 3, combinations of post-decision confidence, decision time, and Remember-Know-Familiar (RKF) judgments were evaluated as postdictors of identification accuracy in a field experiment with ten targets and a very large sample. Fast and confident choosers were highly accurate. Including the RKF judgment did not lead to higher correct classification rates. Participants' self-reported (estimated) decision times also proved to be a postdictor of identification accuracy. Postdicting nonchoosers' identification performance by forming three homogeneous groups of nonchoosers failed, although there were differences with regard to confidence measures and decision times. Experiment 4 tested the usefulness of multiple lineup decisions (portrait face, body, bag, and profile face) for the assessment of identification testimony for nine different targets. Performance in the four different lineup types was not associated with each other, lending support to the idea that multiple lineups can serve as independent sources of evidence. Compared to foil choices and lineup rejections, target/suspect choices were most diagnostic of guilt. The portrait face lineup alone and its combination with the body lineup were most diagnostic for target/suspect choices. To conclude, the present studies suggest that both decision times and post-decision confidence should be collected at the time of identification and be combined in order to assess identification accuracy. Investigators need to be aware though, that there is no postdictive value of nonchooser's estimates. Furthermore, there seem to be no negative effects of target descriptions on identification accuracy when there is a sufficient interval between description and identification, as there is in real cases. Finally, the data speak for the application of multiple lineups with regard to suspect/target choices as a procedure to avoid false identifications, whereas the benefit of multiple lineups for lineup rejections and foil choices seems to be limited. Future studies should address how many and which specific lineup types should be used. In real cases, the results for the assessment (decision times, confidence, decision processes) and control variables (target description, multiple lineups) examined in the present dissertation may vary from those that we obtain in laboratory or field studies. Reasons could be, for example, awareness of the severe consequences of false identifications and false rejections or the stress level at encoding or recognition. It would be interesting to collect data on these issues in real cases so they can be compared to the data obtained in laboratory/field studies. Undoubtedly, it would be a great contribution to the field of identification if data were collected even where DNA samples exist, so that identification accuracy could actually be assessed in real cases.