Background Falls and resulting fractures are a leading cause of morbidity/mortality in the elderly. With the withdrawal of certain selective COX2 inhibitors in 2004, narcotic analgesics have increasingly been recommended as first-line therapy in guidelines for the treatment of chronic pain. Objectives To evaluate the changes in types of medications prescribed for pain pre- and post-withdrawal of certain selective COX2 inhibitors in 2004 and to determine if there was an association with fall events among elderly patients with a diagnosis of osteoarthritis. Design A nested case-control design using electronic medical records compiled between 2001–2009. Setting Electronic medical records for care provided in an integrated health system in rural Pennsylvania over a nine year period (2001–9), the midpoint of which rofecoxib (Vioxx) and valdecoxib (Bextra) were pulled from the market. Participants 13,354 patients, aged 65–89, with a diagnosis of osteoarthritis (OA). Measurements The incidence of falls/fractures was examined in relation to analgesics prescribed: narcotics, COX2 inhibitors, or nonsteroidal anti-inflammatory drugs (NSAIDs). The comparison sample of no fall patients was matched 3:1 to fall patients according to age, gender and comorbidity. Results Narcotic analgesic prescriptions were associated with a significantly increased risk of falls/fractures. The odds ratio of experiencing a fall/fracture was higher in patients prescribed narcotic analgesics than those prescribed a COX2 inhibitor (3.3, 2.5–4.3) or NSAID (4.1, 3.7–4.5). Conclusion The increased use of narcotic analgesics is associated with an increased risk of falls/fractures in elderly patients with osteoarthritis, an observation that suggests the current guidelines for the treatment of pain, which include first-line prescription of narcotics, should be re-evaluated.