The present experiment investigated: (1) the effect of pictorial elaboration on low socioeconomic-status Mexican-American children's long term memory for noun pairs; (2) the effects of labeling mode on estimates of paired-associate acquisition and retention; and (3) the interrelationships between paired-associate task performance and school achievement. The basic design was a 2x2 crossed factorial with pictorial presentation (standard vs. elaborated) and labeling mode (English vs. Spanish). In the acquisition phase of the experiment, 60 second grade children learned a twenty-pair list of common noun referents labeled in English or Spanish. The children learned the list to a leniently scored criterion of 16/20 correct by the study-test paired-associate method. After seven days, retention was assessed by cued-recall and multichoice tests. The results indicated that while elaborated pictorial presentation enhanced the rate of original learning, it neither helped nor hindered the children's long term memory for the pairs. Estimates of original learning and retention did not vary as a function of labeling mode. The rate of paired-associate learning was significantly related only to reading achievement within the elaborated presentation condition. There were no reliable correlations between the indices of paired-associate retention and school achievement. (Author/ AM)