Broadly speaking, we study issues related to metamemory and metacognition: what is the relationship between what individuals know and what they believe they know? To measure this, we ask people to make predictions of future performance, to rate their confidence in past performance, or in some other way provide a subjective assessment of their memory. Then, we correlate these assessments with objective performance on the memory tests themselves. Good metacognitive performance is assessed by high correlations between predicted and actual performance (which rarely occurs, incidentally). More specifically, recent research has focused on several content areas: eyewitness testimony, flashbulb memory, and judgments of learning. We have designed laboratory tasks to present various events to student observers, and we later ask these observers about the events they just witnesses. We've also developed the first set of procedures to study eyewitness memory in product liability cases, as is often the case in civil suits alleging exposure to toxic substances. Over the past decade, we have examined flashbulb memories for a number of significant events: the onset of Desert Storm in 1991, the tragic events on 9/11/01, and most recently the loss of the Space Shuttle Columbia. We not only look at changes memory accuracy, but also in subjective changes in memory. Generally speaking, people do not become more accurate over time, but they do become more confident. Finally, we also use traditional laboratory tasks to study judgments of learning and calibration of comprehension.