Severe health anxiety (SHA) is prevalent, often undiagnosed, persistent untreated, and costly. Meta-analyses have shown effectiveness of mostly individual cognitive behavioral therapy (CBT), whereas Group-CBT has only been studied and shown feasibility in one uncontrolled study of psychiatric outpatients also including self-referrals. We aimed to examine feasibility of Group-CBT for physician-referred psychiatric outpatients with SHA and a future randomized controlled trial (RCT). Group-CBT was conducted in two groups of seven participants. Feasibility was examined about recruitment rate, attrition, organization, and effectiveness. Effectiveness was measured by standardized self-report questionnaires: Whiteley Index-7 (WI-7) and Short Health Anxiety Inventory, and clinician rated remission and functioning from baseline to 3-month follow-up (3FU). Recruitment rate was 1.4 participant per month. Attrition showed dropout of 7%. Organization harbored only one senior therapist. Group-CBT showed effectiveness on all measures except WI-7 at 3FU; this latter explained by participants with concurrent borderline personality disorder. Group-CBT seems feasible for physician-referred psychiatric outpatients with SHA. With modifications, an RCT seems feasible.