Aims: To estimate costs of multiple sclerosis (MS) in a German cohort according to severity of the disease and clinical symptoms. Methods: 144 patients were recruited from an MS outpatient clinic. Costs were calculated according to current German health-economic guidelines from the perspective of the social health insurance system. Patients were either interviewed or completed a questionnaire. Cost assessment covered a 3-month period. Health outcomes were: Expanded Disability Status Scale, MS Functional Composite, Functional Assessment of MS, fatigue, depression (Beck Depression Inventory II) and patients’ socioeconomic status. Multivariate linear regression identified independent cost predictors. Results: Total quarterly costs per patient were EUR 10,329 (95% CI 9,357–11,390). Direct costs were EUR 5,344 for the social health insurance system and EUR 140 for the patient. Drugs represented the major share of direct costs (and 35% of total costs); indirect costs accounted for 47% of total costs. Univariate and multivariate analyses identified age, disability, fatigue and depression as independent predictors for total, indirect or drug costs. Conclusion: MS represents a high economic burden, with direct costs exceeding indirect costs. To reduce costs, research should focus on prevention that slows down progression of MS. Rehabilitation and symptomatic treatment may have merits in decreasing indirect costs.