Background: A commonly used measure to assess mixed features in depressed patients is the Young Mania Rating Scale (YMRS), which only partially aligns with the DSM-5 criteria. Different algorithms on the YMRS have been used to approximate the DSM-5 mixed features criteria. In the present report from the Rhode Island Methods to Improve Diagnostic Assessment and Services (MIDAS) project, we examined the agreement and validity of different approaches towards assessing the mixed features specifier. Methods: Three hundred nine depressed psychiatric patients were interviewed with the Structured Clinical Interview for DSM-IV, the DSM-5 Mixed Features Specifier Interview (DMSI) and the YMRS. Seven definitions of mixed features were examined, two based on the DMSI and five from the YMRS. Results: The prevalence of mixed features varied 8-fold amongst the 7 definitions. The level of agreement between the YMRS definitions and the DMSI was poor. For each definition, mixed features were significantly more common in patients with bipolar disorder than major depressive disorder. A family history of bipolar disorder was significantly associated with the DMSI assessment of mixed features but none of the YMRS approaches. Limitations: The ratings on the measures were not independent of each other. The sample size was too small to compare the patients with bipolar I and bipolar II disorder. Conclusions: While there was evidence of validity for both the DSM-5 and YMRS approaches towards identifying mixed features, the 2 approaches are not interchangeable. The algorithm on the YMRS used to classify patients has a significant impact on prevalence.