Objective The purpose of this study was to estimate and compare the disease burden of depression in Germany using preference-based valuations of depressive health states. Methods The data came from the European Study of the Epidemiology of Mental Disorders (ESEMeD) in which a representative sample of 3555 non-institutionalised adults aged 18 years or older was interviewed in Germany. Computer assisted personal interviews included EQ-51), SF-12, time trade-off (TTO) and contingent valuation (willingness-to-pay, WTP). Respondents in whom 12-month prevalence for a depressive disorder according to DSM-IV was identified, were compared to respondents without depressive disorders (controls). Disease burden was measured in terms of loss of quality-adjusted life years (QALYs) and loss of monetary benefit due to depression per year. Valuations of health-related quality of life (HRQL) were derived from EQ-51) (EQVAS and EQ-5D Index), SF-12 (SF-6D Index) and TTO. Disease burden was estimated by calculating the difference between individual valuation of HRQL/WTP for full health of cases and mean valuation/WTP of controls matched by age and gender. Results were extrapolated to the total population. Results Depending on the valuation method, the loss of QALYs per year due to depression in Germany was 144886 (TTO), 239152 (EQ-5D Index VAS), 265 085 (EQ-5D Index TTO), 307139 (EQVAS) and 403 373 (SF-6D Index). The additional WTP for full health amounted to 4.3 billion Euro. Conclusions Depressive disorders caused a substantial disease burden. Subjects affected by depression showed lower preference-based scores compared to subjects without depression. However, calculated QALY losses varied strongly by valuation method.