Eyewitness testimony can play a crucial role in criminal cases. However, witness memory is fallible. Thus, techniques for assessing witness accuracy are needed. We examined the usefulness of witness recall of contextual information that might be known to the authorities in assessing the likely accuracy of witnesses in describing information unknown and of interest to the authorities. Participants viewed a video of a purse being stolen and were then asked questions about the perpetrator and context of the crime including who (bystanders), when (time of the crime), and where (location of the crime). Participants interrogated soon after witnessing the crime (n=54) exhibited better recall than those not questioned immediately but instead were interrogated for the first time after a 2-day retention interval (n=60). However, interrogation (without feedback) soon after viewing the video reduced forgetting over the 2-day retention interval. Thus, early interrogation can retard subsequent forgetting as suggested by the Testing Effect. Moreover, the quality of recall of the context was positively correlated with recall concerning the perpetrator, and questions concerning bystanders had the highest correlation with perpetrator recall.