The effects of physiological arousal and panic expectancy on anxiety and agoraphobic avoidance were evaluated on thirty-six Ss diagnosed with panic disorder. It was hypothesized that there would be main effects of increased physiological arousal and increased panic expectancy on anxiety and avoidance, as well as an interaction of these two factors, in a behavioral avoidance test containing agoraphobic situations. Physiological arousal induced through hyperventilation raised anxiety immediately following induction, but did not affect anxiety or avoidance on the subsequent walk. Although panic expectancy did not change as a result of the manipulations, initial panic expectancy was the strongest predictor of anxiety on the walk, and the best predictor of general agoraphobia measured by the Chambless Mobility Inventory. Absence of reliable changes in panic expectancy and significant results are discussed in terms of possible pretest sensitization, nature of the sample and subject selection, floor and ceiling effects, experimental demand characteristics, a relatively weak expectancy manipulation, and statistical issues such as large pre-group differences and large within-group variability. This research does however support previous studies which have found a strong correlational relationship between panic expectancy and agoraphobia. / Ph. D.