There is a general agreement that remembering depends not only on the memory processes as such but rather that encoding, storage and retrieval are under the constant influence of the overarching, metacognitive processes. Moreover, many interventions designed to improve memory refer in fact to metacognition. Most attempts to integrate the very different theoretical and experimental approaches in this domain focus on encoding, whereas there is relatively little integration of approaches that focus on retrieval. Therefore, we reviewed the studies that used new ideas to improve memory retrieval due to a “metacognitive intervention”. We concluded that whereas single experimental manipulations were not likely to increase metacognitive ability, more extensive interventions were. We proposed possible theoretical perspectives, namely the Source Monitoring Framework, as a means to integrate the two, so far separate, ways of thinking about the role of metacognition in retrieval: the model of strategic regulation of memory, and the research on appraisals in autobiographical memory. We identified venues for future research which could address, among other issues, integration of these perspectives.