Current controversies concerning outcomes of EEG alpha feedback training are at least in part due to methodological differences among different studies. The aim of this paper is to provide future and present researchers in this field with an analysis of methods used in most of the studies published from 1968 to 1976, and to comment on those methodological issues we think most important. These include alpha assessment, training schedules, and uni- versus bidirectional training. This analysis is accomplished in part by a series of tables listing 45 studies and the detailed methodologies used. It is believed that the probabilities of successful feedback enhancement of alpha can be substantially improved by incorporating the following suggestions: (a) employing at least four training sessions, (b) the use of continuous rather than dichotomous feedback, supplemented by periodic quantitative scores of progress, and (c) using training trials of at least 10 minutes' duration.