Background: DSM-V has been criticized for excessively expanding criteria for bereavement-related depression. The aim of this study was to quantify a potential increase in depression prevalence due to changes in diagnostic criteria and to assess the severity, clinical profile and healthcare use of new cases. Methods: A cross-sectional telephone survey was performed in 2005-2006 in four French regions. Twelve-month prevalence of psychiatric disorders was measured by CIDI-SF. Bereavement was assessed in those who endorsed the gate question to the depression module. Persons with bereavement-related depression according to DSM-IV and DSM-V diagnosis criteria were compared. Results: Of the 22,138 respondents, 692 were bereaved. The prevalence of depression among those bereaved was 49.9% (95% CI ¼=43.7−56.0) according to DSM-IV and 59.6% (53.1−66.1) according to DSM-V [corrected]. The overall prevalence of major depression increased from 8.6% (8.1–9.1) with DSM-IV to 8.8% (8.3−9.3) with DSM-V . Cases diagnosed using DSM-IV presented more symptoms than cases diagnosed using DSM-V but clinical features were similar except regarding criterion E׳s symptoms. Healthcare use was similar between the two groups regarding consultations and psychotropic drug prescription. Limitations: Some DSM-IV and DSM-V criteria were difficult to operationalize in the survey. The observed difference in prevalence according to DSM-IV and DSM-V may be reduced when clinical judgment is taken into account. Conclusions: The overall prevalence of major depression is only marginally increased by the new criteria. However, diagnostic changes increase the prevalence by 10 points among those bereaved. Diagnostic changes do not appear to modify service use.