The purpose of this retrospective, multivariate analysis is to examine how medical conditions and demographic characteristics affect the costs of treating individuals diagnosed with anxiety. Data from MarketScan Databases [The MEDSTAT Group, 2000] were used to identify individuals with new episodes of anxiety. Multivariate analysis was used, with the dependent variable being the log of total medical costs. This analysis controlled for demographic characteristics, medical comorbidities, anxiety diagnosis, and prior resource utilization. A smearing estimate is used to calculate the total medical costs for patients with any anxiety disorder. The mean estimated total medical cost for individuals diagnosed with any anxiety disorder was $6,475. The multivariate model indicates that controlling for demographics and other disease states, generalized anxiety disorder (GAD), panic disorders, and posttraumatic stress disorder (PTSD) are associated with a $2,138, $1,603, and $3,940 increase, respectively, in the total medical cost (P < .0001). The incremental impact of depression, other anxiety disorders, and prior mental health diagnoses on the total medical costs were $1,945, $1,900, and $1,515, respectively (P < .0001). Individuals with the highest costs, and therefore the greatest need for intervention, are anxious patients with depression, individuals diagnosed with PTSD or GAD, and individuals diagnosed with both anxiety and a comorbid medical condition such as an acute myocardial infarction or diabetes. Depression and Anxiety 21:178–184, 2005. © 2005 Wiley-Liss, Inc.