Although panic disorder is classified by the DSM III among the anxiety disorders, there is evidence from epidemiological and neurochemical studies that links panic to the affective disorders. In addition several of the effective pharmacological treatments suggest, as in the depressive disorders, serotonergic involvement. As part of the evaluation of patients who would take part in the Cross. National Collaborative Panic study, polysomnography, the dexametnasone suppression test and platelet imipramine binding and 5HT-uptake were scheduled. 44 patients who met DSM III criteria on the SCID interview schedule for Panic Disorder consented to enter the study. After being medication free for two weeks, these patients had three consecutive all night polyscmnogranis, followed by a dexamethasone suppression test and platelet study of imipramine binding and 5HT-uptake. In our studies we were unable to find for the panic group as a whole a correlation with reported parameters of sleep architecture in endogenous depression. Similarly the imipramine binding differentiated the panickers from the depressed group. Only the 5HT-uptake was similar in both groups. We also found a subgroup of panickers who resembled depressives in terms of the imipramine binding, 5HT-uptake platelet studies and REM latency. In the analysis of the imipramine and 5HT-uptake over the treatment period, we noted changes which suggest a role for serotonin, certainly in a subgroup of panickers and possibly for the group of panic disorder as a whole. © 1990.