A cause-and-effect relation between panic attacks and agoraphobia is an accepted concept. It is believed that, left unchecked, a subgroup of patients with panic attacks will consistently develop agoraphobia. However, to date, there are no means for early identification of this at-risk group. This study analyzed patients with panic attacks and phobic avoidance behaviors by using population-based, survey-collected data. Path analysis was used to determine relations among panic symptoms, phobic behaviors, panic-phobic lag times, and measures of pervasiveness and severity of fears and panic. Panic-related chest pain, dyspnea, trembling, and fear were important factors in the development, pervasiveness, and severity of situational fears and anticipatory anxiety. However, full-blown agoraphobia was only related to the presence of anticipatory anxiety and the pervasiveness of phobic avoidance behaviors. Although the age-of-onset of panic and phobic avoidance was unrelated to other factors, lag times were dependent upon panic symptomatology and the presence of depression. These findings suggest that patients with panic attacks who are at risk for agoraphobia can be identified by the nature of their panic symptoms, and perhaps, through early treatment, the development of phobic avoidance can be averted.