The responses of 6 representative physiological parameters frequently assumed to be measures of anxiety along with a set of 4 psychological tests for measuring anxiety were obtained under naturalistic conditions from 25 patients hospitalized with a first myocardial infarction. A canonical correlational analysis failed to show any relationship between anxiety as assessed by the Taylor Manifest Anxiety Scale, Mood Adjective Check List, State-Trait Anxiety Inventory and Multiple Affect Adjective Check List psychological tests, and anxiety as assessed by the physiological indices of heart rate, systolic and diastolic blood pressures, epinephrine, norepinephrine and VMA. The intercorrelation matrix revealed a significant positive pattern of relationships among all 4 psychological tests, a non-significant, positive pattern of relationships among the physiological indices, and a non-significant, negative pattern of relationships between the psychological and physiological measures. The absence of mood-specific physiological measures for anxiety, as measured by the psychological tests, supports previous theory and investigation and points to the inadvisability of assuming that studies on anxiety that use diverse physiological and psychological measures yield results that may be compared as though they were assessing a common mood.