Studies of eyewitness memory commonly employ variations on a standard misinformation paradigm. Participants are (a) exposed to an event (e.g., a simulated crime), (b) misled about certain details of the event and (c) questioned about their memory of the original event. Misinformation may be provided in the second step via a range of methods. Here, we directly compared the effectiveness of six misinformation delivery methods-leading questions, elaborate leading questions, doctored photographs, simple narratives, scrambled narratives, and missing word narratives. We presented 1182 participants with a video of a simulated robbery and randomly assigned them to receive misinformation about two out of four critical details via one of these methods. In line with the levels of processing account of memory, we report that methods that encourage deeper processing of misinformation result in more memory distortions. Contrary to previous reports, doctored photographs were not a successful method of implanting misinformation. The six delivery methods resulted in minimal differences in confidence and metamemory estimates, but participants were more likely to notice the presence of misinformation in the simple narrative condition. We conclude with suggestions for the selection of an appropriate method of misinformation delivery in future studies. (PsycInfo Database Record (c) 2023 APA, all rights reserved).