1. Five groups of three healthy adult male volunteers (n = 15), all reporting occasional, controlled marijuana use, gave written consent and participated in residential studies lasting 6 to 15 days. 2. Subjects smoked marijuana cigarettes (0, 1.3, 2.3 or 2.7% THC, w/w) at 0945, 1330, 1700 and 2030 every day, and each subject received both active and placebo marijuana cigarettes in 2-5 consecutive day phases, with placebo and active doses presented in an alternating fashion. 3. In comparison with placebo, active marijuana produced a variety of effects on measures of human behavior, including increases in food consumption and errors on psychomotor tasks, decreases in bouts of tobacco-cigarette smoking and verbal interactions and no changes in rates of task performance, time spent under social conditions or social cooperation. 4. Dimensions of human behavior were differentially sensitive to the effects of smoked marijuana. 5. The simultaneous measurement of multiple dimensions of human behavior is a useful procedure for determining dose potency following marijuana administration.