Background: the STOPP criteria advise against the use of long-acting benzodiazepines (LBs). Objective: to study whether LBs are associated with a higher fall risk than short-acting benzodiazepines (SBs) (elimination half-life ≤10 h). Methods: we used base-line data and prospective fall follow-up from the Longitudinal Aging Study Amsterdam, a longitudinal cohort study including 1,509 community-dwelling older persons (Study 1) and from a separate fall prevention study with 564 older persons after a fall (Study 2). Time to the first fall after inclusion and number of falls in the first year after inclusion were the primary endpoints. Results: both in Study 1 and Study 2 the use of SBs was associated with time to the first fall, hazard ratio (HR) 1.62 (95% CI: 1.03-2.56) and HR 1.64 (95% CI: 1.19-2.26),respectively. LBs were not significantly associated with time to first fall, HR 1.40 (0.85-2.31) and HR 1.08 (0.72-1.62). In both studies, the use of SBs was also associated with number of falls, odds ratio (OR) 1.28 (95% CI: 1.01-1.61) and OR 1.37 (95% CI: 1.10-1.70). LBs were not significantly associated with number of falls, OR 1.23 (0.96-1.57) and 1.10 (0.82-1.48). Conclusions: the use of SBs is not associated with a lower fall risk compared with LBs. The use of both SBs and LBs by old persons should be strongly discouraged. © The Author 2013. Published by Oxford University Press on behalf of the British Geriatrics Society. All rights reserved.