Attribute amnesia, a phenomenon in which participants fail to report a just-attended attribute in a surprise test, reflects the importance of expectation in determining memory for attended information. How such expectations arise in the context of attribute amnesia is unclear. The aim of present study was to investigate whether report history can shape expectation, thereby driving attribute amnesia. Participants were initially probed on target location and identity in alternating fashion. They then completed an extended epoch of 360 trials in which only target location was probed, after which target identity was again probed on a critical trial. A substantial performance decrement on the first identity-probe trial that followed the 360 location-probe trials (i.e., the critical trial) was observed, indicating that attribute amnesia can be influenced by recent report history. Although participants were instructed to encode target identity, knew identity could be probed, and had developed familiarity with responding to identity probes, they exhibited a tendency to only encode the target attributes that had been probed on more recent trials. Our results additionally support the robustness of attribute amnesia by replicating the phenomenon under conditions in which the critical probe was familiar and practiced, such that the associated decrement in probe accuracy cannot be attributed to interference from the demands of deciphering an unfamiliar probe question.