Administering lineups "blind"--whereby the administrator does not know the identity of the suspect--is considered part of best practices for lineups. The current study tests whether non-blind lineup administrators would evaluate ambiguous eyewitness statements, and the witness himself or herself, in a manner consistent with their beliefs. College students (n = 219) were told the identity of the suspect or not before administering a lineup to a confederate-witness who made an ambiguous response (e.g., "it could be #3 but I'm not sure"). When ambiguous witness statements matched administrators' beliefs regarding the suspect (compared with when they mismatched administrators' beliefs, or administrators had no belief), administrators (a) were significantly more likely to record the statement as an identification (as opposed to a "not sure" response); (b) were significantly less likely to make statements that might lead the witness away from the suspect; and (c) evaluated the witness's viewing conditions significantly more positively.