Two experiments demonstrate that post-event information, when delivered by another person, can affect people's memory reports. In the first experiment participants were shown several cars, and later, in pairs, given an ‘old’/‘new’ recognition test on these cars plus several lures. There was a small but reliable effect of memory conformity. When the person was given misinformation this lowered accuracy, while presenting accurate information increased accuracy. In the second experiment participants, in pairs, viewed an identical crime except that half saw an accomplice with the thief and half did not. Initial memories were very accurate, but after discussing the crime with the other person in the pair (who saw a slightly different sequence), most pairs conformed. Confidence ratings strongly predicted which person in the pair persuaded the other. Parallels with eyewitness testimony in the Oklahoma bombing case and implications for police interviewing more generally are discussed.