This study assessed the effects of bizarreness, prompt modality, and prompt type for 144 five and eight year-old children on recognition memory of pictorial pairs. Presentation of stimuli was self-paced, allowing for the collection of study time and response latency data, as well as recording number correct. While both bizarre and nonbizarre forms of elaboration significantly facilitated performance relative to a standard condition, bizarre pair depictions significantly depressed noun-pair learning for five year olds. A reliable difference between bizarre and nonbizarre pairs for eight year olds was not observed. In addition, the combination of nonbizarre pairs and verbal prompts led to increased performance for five year olds, whereas nonbizarre pairs and pictorial prompts led to increased performance for eight year olds. With bizarre pairs, there was no variability attributable to modality for either age group. (Author)