Casner (2006) tested the ability of a prosopagnosic, LB, to discriminate objects that would require the use of the coordinate relations recognition system posited by Cooper and Wojan. Casner reported that the prosopagnosic, LB, was impaired in coordinate relations tasks, but did not differ significantly from controls in all tasks that only required categorical recognition. However, Farah, Levinson, and Klein (1995) reported on a patient who was not impaired at discriminating eyeglasses, a task for which the coordinate relations hypothesis would predict an impairment. The present study replicated Farah et al.'s paradigm with the prosopagnosic, LB. One experiment found that LB, relative to controls, was significantly impaired when discriminating eyeglasses requiring the use of a coordinate relations recognition system rather than a categorical recognition system, however, the other experiment failed to find such a difference. Two experiments results may have arisen from a strategy that LB developed used.