In this paper the authors examine the interrelationship of both the noradrenergic (NA) system and the hypothalamic-pituitary-adrenal (HPA) axis and its implications for panic disorder (PD). Seventeen PD patients and 16 healthy volunteers were challenged orally 12 weeks apart with the alpha 2-agonist clonidine (13 healthy volunteers and 12 patients repeated the challenge). Between challenges, PD patients were treated with fluoxetine, with 10 of 12 improving at least moderately. Both during the acute phase of the illness and during the phase of pharmacological improvement, patients demonstrated a greater percentage of reductions of plasma 3-methoxy-4-hydroxyphenylglycol (MHPG) and plasma cortisol during clonidine challenge. We used correlational matrices to examine the relationship between the NA system, as reflected by plasma MHPG, and the HPA axis, as reflected by plasma cortisol measures. Healthy volunteers exhibited multiple significant "couplings" between either baseline or maximal decrease (delta max) of plasma MHPG, with either baseline or delta max plasma cortisol measures both within the first and second challenges and between the first and second challenges. In contrast, PD patients demonstrated "uncoupling" of the NA system and the HPA axis, with no significant correlations observed between either baseline and/or maximal decrease (delta max) measures of MHPG with the same cortisol measures for either the first or second challenge. The same uncoupling was observed for NA/HPA correlations between the first and second challenges. These data suggest that the hyperresponsivity to clonidine in PD patients persists during fluoxetine treatment.(ABSTRACT TRUNCATED AT 250 WORDS)