In this study, temporal speech cues were integrated into online curriculum to solve non-routine problems in curricular multimedia. Teachers-in-training (n=56) were randomly assigned to one of three treatment conditions. It was expected that participants in the temporal speech cues condition would be more likely to solve problems than those in the expository text, or the narrated text conditions. The study used a 3-factor repeated measures design, a pre-test, a post-test, and a delayed post-test. The delayed post-test was included to determine the content of long-term memory over time. Results of the study confirmed the expectation that, after a latency period, participants learning from temporal speech cues applied more instructional design solutions to non-routine problems than those learning from expository text, and narrated text.