Previous research on eyewitness psychology has failed to predict who is likely to provide good eyewitness accounts in terms of either person identification or verbal recall. The current article presents two studies that explore two potential correlates of eyewitness facial identification performance: (1) perceived objectivity as an eyewitness and (2) observers’ attitudes toward eyewitness performance. Also explored in these studies was whether facial identification performance could be predicted from observers’ confidence in their abilities as eyewitnesses. Study 1 involved 96 undergraduate student participants who completed a perceived objectivity scale and an eyewitness accuracy scale before viewing a 3-minute color video depicting a purse snatching incident. Following the viewing, participants identified the suspect from a group of photographs. Study 2 involved 160 undergraduate student participants who completed the same scales and viewed the same purse snatching incident as the participants in study 1. Following the viewing, participants were given a blank paper and asked to provide a verbal description of the suspect. Results of both experiments indicated that none of the variables under consideration as predictors of eyewitness performance were significantly related to identification performance. However, participants who reported more objectivity and confidence in their eyewitness performance were no more likely to provide accurate suspect descriptions and in fact included more incorrect elements in their verbal descriptions. The findings indicate that judges should take more care in instructing jurors regarding eyewitness testimony. Tables, references