Free recall, cued recall, and clustering performance of fourth grade lower socioeconomic black and white children is studied under four conditions of teaching: random presentation, random presentation plus instruction, blocked presentation, and blocked presentation with instruction of categorically related list. Results indicate that both recall and clustering performance is enhanced by task structure in the form of blocked and blocked plus instruction presentation. Contrary to expectations, color differences in recall exist. The effect of blocking categorized items is especially facilitative for black children. Free recall is significantly better for black children who receive the categorized items contiguously than for black children who are presented the same items randomly. Caution is warranted when investigating cognitive development in lower SES children. Correlational analysis of the use of clustering during free recall reveal no significant relationships and seriously questions the use of an experimenter defined organizational strategy to assess level 2 ability as proposed by Jensen. Level 2 cognitive ability such as the recall of categorically related items can be used when teaching conditions encourage their use. Even then, however, a production deficiency may be present that does not permit a clear picture of the full extent of development. (Author/AM)