Objective: As older adults become increasingly reliant on emergency departments (EDs) for care, there is an interest in determining what types of ED visits by this population may be preventable, or amenable to other forms of care. The aim of this project was to explore the concept of preventable ED visits by older adults. Methods: We conducted a literature search to identify definitions of "preventable" or "avoidable" ED visits. We then applied a definition of preventable ED visits to an administrative data set consisting of ED visit data extracted from four sites in Halifax, Nova Scotia, Canada. Visits for patients 65 years of age or older were eligible for inclusion. Visits were categorized using triage level and discharge diagnosis. Results: Four methods of defining preventable ED visits were identified in our literature search: 1) Ambulatory Care Sensitive Conditions (ACSCs) (N = 7), 2) Low Acuity/low intensity visits (N = 5), 3) New York University (NYU) (Billings) Algorithm (N = 3) and 4) hospital admission vs. non-admission (N = 1). We categorized 34,454 ED visits from our dataset using a modified definition of preventable ED visits that included ACSCs (15.3%) as well as low acuity visits that required no testing or hospital admission (9.9%). Conclusions: Our results suggest that approximately 25% of ED visits by older adults may be preventable or amenable to other forms of care. This data may be useful in the planning of care delivery appropriate for the needs of this population.