BACKGROUND: Explicit criteria for determining potentially inappropriate medication consumption in elderly were elaborated by Beers et al. These lists have been used worldwide to evaluate medical prescriptions but there is little epidemiologic evidence demonstrating negative consequences of inappropriate medication use. It has been reported that some drugs could increase the risk of falls, which are a frequent and serious problem in elderly population. We aimed to evaluate the association between the use of potentially inappropriate medications and the risk of falls. METHODS: The 3C Study is a multicentre prospective cohort study conducted in France with 4 years of follow-up. Non-institutionalized men and women aged 65 years or over (N = 6343) were randomly selected from electoral rolls. Data on socio-demographic, medical characteristics and medication use (based on self-reports and data from the national healthcare insurance) were collected. Use of inappropriate medication for elderly was defined from established criteria. Data about falls were collected at the two follow-up examinations (2 years and 4 years after baseline). The association between the exposure to inappropriate medications and the risk of falls was evaluated using multivariate models (Cox model and logistic regression). RESULTS: 32% of subjects reported inappropriate medication use at baseline and 29% at least two of the three examinations; 22% had fallen 2 times or more during follow-up. Overall, inappropriate medication users had an increased risk of falling. This increase was mainly due to the use of long-acting benzodiazepines (adjusted odds ratio (OR) = 1.4, 95% confidence interval: [1.1-1.8], in both occasional and regular users), other inappropriate psychotropics (adjusted OR = 1.7 [1.7-2.7] in regular users), or medication with anticholinergic properties (adjusted OR = 1.6 [1.2-2.1] in regular users). Neither occasional, nor regular use of short- or intermediate-acting benzodiazepines was associated with an increased risk of falling. Further analysis in long-acting benzodiazepines users did not show any dose-effect relation between the number of prescriptions filled over a 3-year period and the risk of falling. CONCLUSION: Our study showed that use of inappropriate medications was associated with an increased risk of falling in elderly persons. This increase was mainly due to long-acting benzodiazepines and other inappropriate psychotropics, and to medications with anticholinergic properties.