Two hundred consecutive patients with femoral neck fractures (FNF) and 200 controls matched by age, sex, nursing home residency and number of admissions within the last 2 years, were interviewed about recent drug use. The matching criteria were chosen to achieve case and control groups that were comparable with regard to overall physical disability. Thus a preliminary study had shown FNF to be strongly associated with nursing home residency and to be correlated with the number of recent hospital admissions. FNF was found to be negatively associated with use of coronary drugs (OR = 0.27; 95% CI = 0.11-0.70) and positively associated with drug use in general (OR = 2.00; 95% CI = 1.04-3.05). The OR for the various classes of psychotropic drugs was consistently greater than 1., although this was not statistically significant. An inherent problem in interpreting such findings is that underlying disease may predispose the subject to both FNF and drug use, thus creating a non-causal relationship.