A pool of 200 males, aged 18-22 and enrolled in introductory psychology at Purdue University, were screened on the Eysenck Personality Inventory - Neuroticism Scale and the Marlowe-Crowne Social Desirability Scale. Forty of these males, falling within one standard deviation of the norm on both screening measures, were assigned to one of four groups--recorded control (RC), induced affect (IA), induced positive affect (IPA), or induced negative affect (INA). Subjects in each group participated in three sessions, one session during which experimental rationale was explained and informed consent obtained and two experimental sessions. During both experimental sessions, each subject was recorded for electroencephalogram (EEG) and electrocardiogram (EKG) during 20 minutes of baseline and 30 minutes (five six-minute phases) of experimental intervention. During baseline and experimental phases, all subjects reclined on a hospital bed, the head of which was in a semi-upright position, with their eyes gently closed. They were asked to maintain a state of relaxed alertness during baseline. During the five six-minute phases of experimental intervention in each session, subjects in group RC were read the Constitution by the experimenter, while subjects in the experimental groups (IA, IPA, and INA) underwent two phases of induction of emotional arousal, alternated with three phases of relaxation. During arousal phases, the experimenter facilitated subjects' self-induction of arousal through use of standard induced affect instructions. Because investigation of physiological concomitants of positive and negative emotions was of interest, subjects in one experimental group received instructions for induction of positive affect while subjects in another experimental group received instructions for induction of negative affect. A third experimental group received standard induced (non-specific) affect instructions, which allowed them to develop either positive or negative emotions. Theta (3-7 cps) and alpha (8-13 cps) activity was filtered electronically from raw EEG activity as it was recorded on paper by a Lafayette Instruments polygraph, and seconds per minute of occurrence of each was counted by electronic counters and recorded from a visual display by the polygraph operator. EKG was recorded on paper simultaneously with EEG and was counted manually after experimentation by the experimenter and one other rater. At the end of each phase, baseline and experimental, each subject completed a state measure of affective arousal, the Affect Rating Scale - Now, which consists of five components--potent positive, potent negative, weak negative, weak positive, and relax. Results obtained indicated a tendency for subjects undergoing affective arousal, negative or positive, to produce increased theta EEG activity, relative to baseline and alternated relaxation phases, concomitant with elevated heart-rate and self report of strong emotional arousal. The strongest effects were obtained for subjects in group INA. Results obtained with alpha EEG activity were less clear but suggestive of increased alpha activity during the final arousal phase relative to the preceding relaxation phase in the final experimental session for both positive and negative arousal. No strong consistent progressive change over relaxation or arousal phases was observed for physiological or self-report indices. Results obtained do not support activation theories of emotional arousal but do offer tentative support for Lacey's contention that EEG synchronization following cardiovascular system activation provides a mechanism for organismic physiological homeostasis. Also, the results of this study offer tentative support for Schachter's theory that the cognitive component of emotional arousal is largely responsible for discrimination of positive and negative emotional states. More detailed investigation is necessary before more substantive conclusions can be drawn.