A number of naturalistic studies have found that medical co-morbidity conveys a worse long-term prognosis in the treatment of major depressive disorder (MDD). The purpose of this study was to test whether the presence of co-morbid medical conditions can predict clinical response in patients with treatment-resistant MDD (TRD) treated with open-label nortriptyline (NT). Ninety-two patients with TRD entered a 6-week open trial of NT. The presence of co-morbid medical disorders was assessed during the screen visit. The degree of medical co-morbidity during the screen visit was then quantified with the use of the Cumulative Illness Rating Scale-Geriatric Version (CIRS(G)). We tested whether CIRS(G) scores predicted clinical response or depression severity at endpoint. CIRS scores at baseline did not significantly predict treatment response. The results of this study fail to confirm the relationship between co-morbid medical conditions and poor outcome in the treatment of MDD for patients with TRD. Patients with TRD and co-morbid medical conditions can be expected to respond to antidepressants as well as their counterparts without concurrent axis III co-morbidity. The CIRS(G) scores for this TRD sample were lower than those reported for geriatric depression, or for depressed patients with severe medical illness, common in medical and surgical wards and in most specialty clinics of large academic centers. Thus, the present results cannot be generalized to such populations.