As an extension of an earlier investigation that examined the effects of mnemonic strategy application on children's memory for abstract prose passages, a study compared the benefits accrued by students taught two different variations of the mnemonic keyword method for learning abstract prose information, via tasks of associative recognition and associative recall, both immediately after instruction and three days later. Subjects included 183 eighth-grade students assigned to one of four instructional conditions: (1) keyword-symbol picture, which provided semantically based representations and conventional symbols for encoding the relatively abstract attribute information and then integrated picture referents with a keyword referent; (2) keyword-keyword picture, which provided phonetically based concrete words for encoding the relatively abstract attribute information and then integrated the picture referents with a picture of another keyword referent; (3) control-passage, which provided students with a second opportunity to read the story passages and its associated attributes; and (4) control-list, which provided students with a list of each name with its associated attributes. Results supported the efficacy of semantically and phonetically based strategy applications for learning abstract prose information. (Appendixes include story passages, instructions, study passages, and test pages for all conditions.) (HOD)