In forensic contexts it is common that witnesses retell and discuss the experienced event many times. It is of forensic importance to understand how this influences memory and meta-memory. Eighty-nine participants viewed a short film and were assigned to one of four conditions: (1) Laboratory discussion (five discussions of the event with a confederate), (2) Family discussion (five discussions of the event with a family member), (3) Retell (five retellings of the event) and (4) Control. Three weeks later participants gave an open free recall, and then 3 days later confidence judged the recalled information. The results showed significant differences between the four conditions on number of correct items, incorrect items, accuracy, confidence and calibration. The results suggest that discussion of an experienced event may reduce some of the beneficial memory and meta-memory effects caused by mere retelling, but may have no great negative effects compared to a control condition. Copyright © 2010 John Wiley & Sons, Ltd.