Two prevailing models for teaching memory strategies for remembering were compared for two kindergarten classes (45 boys, 30 girls). A third class provided a control. One class received instruction in specific and efficacious strategies for recall on three discrete tasks: sequential memory, paired-associate memory, and clustering recall. The second class received the same instruction embedded in an overt verbalization procedure adapted form Bash and Camp (1985). The control class received as much practice as the two instructional groups but no instruction on strategy or overt verbalization. Significant gains in memory were made on paired associates at posttest for both the strategy-only group and the overt-verbalization group. Overt verbalization inhibits recall on paired associates at delayed posttest and its effect on sequential memory at posttest remains equivocal.