Background: Medical residents frequently participate in code status discussions, yet their competency in this role has not been evaluated. Purpose: The objective is to determine the quality of code status discussions from the perspective of both resident and patient. Methods: We conducted a cross-sectional survey of consecutive pairs of hospitalized patients and admitting residents at Mayo Clinic in March 2007. We measured perceptions of occurrence and content of code status discussions, admission volume, and demographic information. Results: Among the 41 matched pairs, residents and patients agreed that a code status conversation occurred in 63% of cases. Agreement was more likely if residents performed less than 4 admissions (p= .02). Patients reported the inclusion of specific discussion components, such as resuscitation procedures (7%) and outcomes (0%), less frequently than residents (71% and 27%, respectively, p< .001). Conclusions: Residents and patients demonstrated poor agreement on the occurrence and components of code status conversations. Residency programs should identify ways to enhance residents' competency in eliciting patients' code status preferences and provide adequate time for code status discussions. [ABSTRACT FROM AUTHOR]