This paper outlines and analyzes the important methodological and experimental design considerations uncovered by recent inquiries into the effects of alcohol on human behavior. The major issues discussed include: (1) the need and technique for complete manipulation of subjects' beverage expectancies in conjunction with alcohol and placebo administrations; (2) subject selection, with particular emphasis on the importance of subjects' prior experience with alcohol and how this might be assessed; (3) dosage, type, and measurement factors in alcohol administration; and (4) subjects' predispositions, the nature of their experimental tasks and the importance of the setting and social context of the experiment. (Author)