This study begins a systematic examination of the role that content of materials used in experimental free recall tasks, as well as the origin plus the appropriateness of designated "conceptual categories," play in determining mnemonic performance of different groups. It will deliberately engineer the way materials are derived and constructed for recall tasks. This study started by eliciting information on a set of categories generated by those who have been consistently shown to cluster and recall relatively poorly--lower class black high school dropouts. They were compared with white middle class parochial high school students on tasks developed from experiences of similar urban black youth. Prior to conducting the recall study, a group of 75 black adolescents living in and attending school in Central Brooklyn were employed to generate recall materials. Thirty-four black and white adolescents recalled a list of 30 words. There was no significant difference between groups in the amount recalled. Black adolescents clustered more than white adolescents with a high correlation between amount recalled and clustering. Analysis of within list category responses suggests that the "black categories" depressed the categorized recall of white adolescents. The results indicate that cultural origin of materials can alter standard expectations of ethnic group performance. (Author/JM)