The effect of different textual versions (macroscopic (control), submicroscopic, and guided imagery) of the explanation of a chemical phenomenon on students' submicroscopic explanation of a related phenomenon was examined. The sample included 152 pre-service science teachers. The three textual versions of the explanation were distributed randomly to the participants. The results revealed that students who received the submicroscopic version and those who received the guided imagery version outperformed students who received the macroscopic version. These results indicated that students' use of the submicroscopic understanding was not spontaneous and they needed to be cued to do so. Also, the submicroscopic mean scores of all three groups were low, and this might be an indication of a weak ability to: 1) translate between macroscopic and submicroscopic levels of chemistry, and 2) transfer a submicroscopic understanding of one phenomenon to another related one. (Contains 6 tables and 1 figure.)