Studies on the association between long-term benzodiazepine use and brain abnormalities have yielded conflicting results. The computed tomographic (CT) scans of 20 long-term users of benzodiazepine (65% men; mean age +/- SD [range], 42 +/- 12.1 years [23-59]; mean daily benzodiazepine dose [diazepam equivalents], 19.5 +/- 16.2 mg [2.5-70]; mean cumulative benzodiazepine exposure, 55.2 g [1.8-198]) were compared with 36 age- (+/-3 years) and sex-matched controls. Controls were prospectively recruited from 96 patients attending a neurology clinic and were interviewed to screen for alcohol and substance use disorders and other conditions possibly leading to brain atrophy. Three neuroradiologists blindly assessed each CT scan for atrophy and measured ventricles (V1, V2, V3), sulci, fissures, cisterns, and folia. Reliability among observers ranged from 0.92 to <0.1, in which case deleting one observer increased all reliabilities to >0.45. No difference in atrophy was found between benzodiazepine users and controls. V1 measures were significantly higher for benzodiazepine users than for controls (mean +/- SD, 12.1 +/- 1.3 vs. 11.1 +/- 2.0;p = 0.02), but measures of third and fourth largest sulci were significantly higher in controls than in benzodiazepine users. Right third and fourth largest sulci (mean +/- SD), respectively, were the following: controls, 0.72 +/- 0.4 and 0.74 +/- 0.7; benzodiazepine users, 0.51 +/- 0.3 and 0.46 +/- 0.3 (p < 0.02). Left third and fourth largest sulci, respectively, were the following: controls, 0.77 +/- 0.6 and 0.65 +/-0.3; benzodiazepine users, 0.53 +/- 0.3 and 0.5 +/- 0.3 (p < 0.02). Long-term benzodiazepine therapy does not result in brain abnormalities that can be demonstrated on CT scans.