Self-reported dietary intake is commonly used to inform policy; however, memory-based reports are subject to error. Our aim was to examine dietary reporting errors using a repeated-events framework. Participants (N = 102) completed a 3-day food diary and 10 days later recalled what they had consumed on one self-nominated day and one experimenter-nominated day from the diary period. Self-nominated day reports were more accurate than experimenter-nominated day reports. Across both days, participants made more errors by reporting a food from the wrong day than by reporting foods not recorded in the diary at all. Unexpectedly, participants who completed their food-diary across Sunday-Monday-Tuesday were more accurate than those who completed across Thursday-Friday-Saturday, and participants who completed the study in 2020 were more accurate than those who completed it in 2021/2. Overall, results are consistent with the repeated events literature and outline a new approach to better understand dietary self-reporting.