Studies of schizophrenia suggest a specific impairment in binding different parts of a memory event into a cohesive whole, a finding that may account for the reported preferential deficits in associative recognition memory relative to item recognition. As a further test of this hypothesis and to exert greater control over task differences, we used a recognition memory interference test in which participants encoded landscape pictures that had each been divided into three segments. During encoding, subjects were presented with one segment from each of the landscapes. Then, an interference generating task followed consisting of the presentation of the second segment from half of the landscapes. Finally, a forced-choice recognition memory test consisted of the presentation of each encoding picture stimulus concurrently with the related third segment that had never been presented before. Thus, for half of the stimuli, additional related information was encoded and this is known to interfere with recognition memory. However, an impaired ability to bind this related information should reduce the interfering effect of associated stimuli. Thirty-five schizophrenia patients and 37 healthy controls were administered this memory interference task. A significant interaction between groups and recognition conditions was found with a significant interference effect observed for controls (performance dropping from 76% to 62%) but not for patients (performance remaining unchanged from 66% to 68%). These results provide further support for faulty associative memory processing in schizophrenia.