The role of quality measurement of medication use is becoming more critical as consumers, employers and others demand increased accountability and transparency from the delivery of health care. At this time, however, there are no explicit quality indicators of preventable drug-related morbidity (PDRM) that could be used by clinicians and/or a health care organization. OBJECTIVES: To create consensus-approved clinical indicators of PDRM in older adults applicable to the Canadian health care system. METHODS: A written survey was constructed, listing the clinical outcome and pattern of care related to a number of possible PDRMs in older adults. Using the Delphi technique, two independent six-member expert panels (geriatricians, clinical pharmacologists) in Nova Scotia, Canada were asked to judge whether the outcome in each situation was foreseeable and recognizable, and whether causality was identifiable and controllable. The panel could also suggest additional PDRMs. Subsequently, a focus group of 12 general practitioners (GPs) evaluated these PDRM indicators. The inclusion of this third panel provided a triangulation of expert opinion across three practice areas. RESULTS: The two expert panels proposed 58 indicators of PDRMs in older adults after two rounds of the Delphi technique. The GPs agreed with 52 (90%) of these PDRM indicators. CONCLUSIONS: This study showed that consensus on quality indicators of PDRM can be reached among experts. These indicators could be used by a health care organization to proactively identify patients at risk for a PDRM and to improve the quality, safety and appropriateness of medication use. Additionally, the indicators form an important bridge between processes and outcomes of care and could be used in conjunction with existing medication use quality indicators. Subsequent phases of this study will involve pharmacist validation of these PDRMs, and identification (through an integrated medical database of older Nova Scotians) of patients who experienced PDRMs.