Sixty-four learning disabled sixth, seventh, and eighth grade students were assigned to 24 small instructional groups which were in turn assigned at random to mnemonic and control conditions. Instructional groups were then given three daily lessons on dinosaurs in counterbalanced order. The lessons involved dinosaur vocabulary, attributes of dinosaurs, and reasons for dinosaur extinction. Mnemonic groups were taught the information using keyword and pegword techniques. Control instructional groups were taught the same information using the principles of direct instruction, including teacher-directed questioning, choral group responding, fast instructional pacing, and cumulative review. A test was given after each lesson on that lesson's content. On the fourth day, all students were given production and identification tests on the content of all three lessons. Results indicated that mnemonic groups outperformed control groups on both delayed tests as well as the attributes and extinction lesson. Mnemonic groups also outperformed control groups on the vocabulary lesson and on a test of applicaton of dinosaur word parts. In addition, it was found that control subjects' responses revealed significantly more intralist intrusions than mnemonic subjects'. No meaningful trend across days of instruction was observed for either condition, and interference across lessons (interlist intrusions) were not observed in either condition. Results suggested that (1) previously validated mnemonic techniques were successful over several days of instruction; (2) mnemonic instructional groups performed significantly better than a direct questioning condition on a delayed test of extended instruction; (3) mnemonic and control Ss were equally able to apply learned vocabulary information; and (4) mnemonic instruction appeared to exhibit no inhibiting effect over several different lessons and days of instruction. (Author/CL)