To the Editor.—Ìý In response to the letter by Leckman et al that appeared in the July Archives,¹ we agree that the differences in results probably reflect differences in the methods of these studies. First, they are fundamentally different studies, as Dr Leckman pointed out. Ours were studies of patients with primary anxiety disorders^2,3 and theirs was a study of primary affective disorder,⁴ so the difference in results may simply indicate that patients with primary depressive disorder who have panic attacks are not the same as patients with primary anxiety disorder. We doubt that differences in the selection of our controls account for the difference in the study results—while three of our control probands had major depressive disorder, only two (0.8%) of the relatives in their study were so diagnosed.Dr Leckman's final two points dealt with the handling of primary and secondary cases of both anxiety