Background and aims: Reports of diseases and health problems vary as a function of the information source. In the present study we compared the capture of information on morbidity using medical records, self-reports, and drug registrations. Methods: A concurrent review of medical records, self-reports, and registration of marker drugs was conducted to determine diagnosis for 44 common diseases. Diagnoses from the various sources were uniformly classified according to ICD-10. The study included a sample of 702 individuals, aged 80 and older, enrolled in population-based longitudinal studies. Results: The morbidity rates differed considerably across the used sources. Although medical records captured most of the morbidity, self-reports offered supplemental information especially for less objective health problems. Marker drugs typically confirmed information in the records, but only for a limited number of diseases. Discussion: In studies of aging and health, a thorough review of medical records and a concurrent evaluation of self-reports and marker drugs represent a valuable strategy for portraying morbidity. This strategy goes beyond the use of a single source like self-reports, and provides better estimates of health conditions in the elderly.