This study explores the reliability of a data source on the quality and content of care rarely used in studies comparing the performance of health care organizations, that is, patient reports obtained from surveys. Evidence of patient survey reliability and validity and report data on patient reporting accuracy were reviewed for ten events that may have occurred during an initial health assessment for new adult enrollees of a health maintenance organization (HMO). Reports of 380 patients obtained through telephone survey were compared with medical records. For chest radiograph, mammogram, and electrocardiogram (EKG), patient reports exhibited both sensitivity and specificity. For serum cholesterol test, patients proved to be sensitive but not specific reporters. For blood pressure measurement, stool kit, and rectal examination, false negative rates were low (less than or equal to 0.10); they were somewhat higher for breast self-examination instruction and pelvic examination (0.21 and 0.22, respectively). Only for testicular self-examination instruction did patient reports fail to confirm medical record documentation (false negative rate = 0.53). Multivariate analysis showed a small association between increasing patient age and decreasing confirmation. Gender did not affect reporting ability, and agreement did not deteriorate over a 2- to 3-month postencounter interval. Patient reports appear to merit greater use in comparative studies of technical quality of care. Key words: quality of health care; quality assurance; health care; ambulatory care; patient recall; patient reports.