The use of benzodiazepines (BZD) or other psychotropic agents is an established risk factor for falls and hip fractures. The evidence supporting this association has been based solely on history and/or prescription data. In a case-control study we monitored the intake of BZD, tricyclic antidepressants (TCA) and barbiturates in patients with hip fracture by serum analysis, and compared this measurement with drug history and prescription records. The serum was BZD positive in 41% of 82 community-dwelling patients; an equivalent by history and/or prescription records, however, was detected in only 18% (p < 0.0001). In contrast, in 82 age- and gender-matched community-dwelling control patients, there was only a minor discrepancy (3%) in BZD use between anamnestic (N = 19) and analytical (N = 21) data. In 23 patients admitted from nursing homes, a similar trend was visible (39% serum positive for BZD vs 26% by drug history). Barbiturate use (N = 2) could be neglected, and TCA intake was minor (N = 7); in each case both assessments were in accordance. In conclusion, in patients with hip fracture, BZD use is substantially more frequent than reported and previously assumed, pointing to an underestimated risk factor for this injury.