Using mathemagenic-based research as a guide, this study attempted to determine whether the location of a picture (pre- or post-reading passage) or the type of picture (overview or specific incident) could affect reading comprehension. Ninety-two college students were randomly assigned to one of the cells in a two-by-two factorial design and required to read a text-like passage of approximately 1200 words. Results add support to the idea that pictorial attributes can serve mathemagenic functions. As with "inserted questions" research, this study found location to be a significant instructional variable, since pictures placed after a reading passage increased learning more than pictures placed before the reading passage. A list of references is provided. (Author/RAO)