A group of preservice science teachers edited video footage of their practice teaching to identify and isolate critical incidents. They then wrote guided reflection papers on those critical incidents using different forms of media prompts while they wrote. The authors used a counterbalanced research design to compare the quality of writing that participants produced when they had access to either their edited video clip of the incident, audio from the clip only, or their memory of the incident alone while writing. All reflection papers were evaluated using a rubric developed by Ward and McCotter (2004). An analysis of variance among paper scores showed that participants wrote significantly higher quality papers on several indicators when prompted by video than when prompted by audio. There was also a difference in means between their reflections when prompted by video and when they worked from memory alone.