Aim Short term administration of benzodiazepines (BZD) was found to prolong reaction time (RT) in experimental studies. However, studies on long term BZD use did not always adjust for important confounders and showed inconsistent results. We aimed to identify a possible relationship between long term BZD use and RT in BZD users in this large cross-sectional, observational study. Methods The RTs of non-users (n = 2404) were compared with low (n = 288), intermediate (n = 74), and high dose BZD users (n = 57) in the Netherlands Study of Depression and Anxiety (NESDA). RTs were obtained from the Implicit Association Test. Analyses were adjusted for sociodemographic characteristics, health indicators, severity of psychopathology and antidepressant use. Results Of the NESDA participants, 419 subjects (14.8%) used BZDs. A higher dose of BZDs was associated with prolonged RTs (P = 0.01). When comparing the different dose groups, the high dose group, but not the low and medium dose groups, had significantly longer RTs than the non-users. Conclusions Tolerance for the RT prolonging effect of relatively high doses of BZDs does not seem to develop. As prolonged RTs can have adverse consequences in daily life, BZDs should be prescribed conservatively at the lowest possible dose. © 2013 The British Pharmacological Society.