Use of standardized patients in training research staff has been limited. This study evaluated its feasibility. An expert panel created six scenarios using standardized patients to portray dyads of dementia patients/caregivers, plus instructions for actors. Three research assistants trained in administering the Hamilton Depression Rating Scale portion of the Structured Interview for the Clinical Assessment of Depression in Dementia administered it to each dyad. An expert panel member telemonitored each session and scored pairs using the same instrument. Sessions were videotaped, watched and scored by research assistants. Their scores were compared with expert ratings, and deviation scores were calculated, with mean item deviation scores compared using analysis of variance. Interclass correlations and analysis of variance revealed no differences between research-assistant ratings and their ratings compared with a standard, supporting the feasibility of using standardized patients to train research assistants to perform complex clinical assessments.