In six studies, Event-Related Potentials (ERPs) were employed to investigate the neural basis of conscious (explicit) retrieval from long term memory. The studies provide the first detailed reports of ERP effects associated with explicit retrieval on tests of word-stem (e.g. TRE) cued recall and stem completion. The relationship between the cued recall ERP effects and those previously observed on other tests of memory was also investigated. This was done by directly contrasting ERP correlates of explicit retrieval on the cued recall and recognition memory tasks. Two features of the cued recall and recognition memory ERP effects were found to be highly similar: ERP effects for each task were comprised of parietally and frontally distributed components which differed, as a function of task, only in two respects. First, the parietal effect for cued recall appeared to be somewhat delayed in onset latency relative to that for recognition memory. Second, the hemispheric asymmetry of the frontal effect for cued recall was less marked than that for recognition memory. The two ERP components were interpreted in terms of processes contributing to the recollection of previous episodes in which words were presented for study. A basic distinction between retrieval and post-retrieval processes was invoked to provide a functional account for the two components. The parietal component was related to retrieval processing associated with 'old/new' judgements. The frontal component was related to post-retrieval processing of retrieved information, which may be more under strategic control, and therefore sensitive to factors extrinsic to those affecting retrieval success per se. In conjunction with the findings of other ERP studies of long term memory, the present results suggest that similar electrophysiological and cognitive processes may be involved in retrieval and post-retrieval processing on a wide range of memory tasks. It is proposed that, under certain conditions, a common feature of these tasks may be the requirement to engage working memory, to monitor explicit retrieval processing.