Background: Studies assessing the prevalence of depression and anxiety in multiple sclerosis (MS) have used various ascertainment methods that capture different constructs. The relationships between these methods are incompletely understood. Psychiatric comorbidity is associated with lower health-related quality of life (HRQOL) in MS, but the effects of past diagnoses of depression and anxiety on HRQOL are largely unknown. We compared the prevalence of depression and anxiety in persons with MS using administrative data, self-reported physician diagnoses, and symptom-based measures and compared characteristics of persons classified as depressed or anxious by each method. We evaluated whether HRQOL was most affected by previous diagnoses of depression or anxiety or by current symptoms. Methods: We linked clinical and administrative data for 859 participants with MS. HRQOL was measured by the Health Utilities Index Mark 3. We classified participants as depressed or anxious using administrative data, self-reported physician diagnoses, and the Hospital Anxiety and Depression Scale. Multivariable linear regression examined whether diagnosed depression or anxiety affected HRQOL after accounting for current symptoms. Results: Lifetime prevalence estimates for depression were approximately 30% regardless of methods used, but 35.8% with current depressive symptoms were not captured by either administrative data or selfreported diagnoses. Prevalence estimates of anxiety ranged from 11% to 19%, but 65.6% with current anxiety were not captured by either administrative data or self-reported diagnoses. Previous diagnoses did not decrease HRQOL after accounting for current symptoms. Conclusions: Depression and, to a greater extent, anxiety remain underdiagnosed and undertreated in MS; both substantially contribute to reduced HRQOL in MS.