The objective of our study was to investigate whether use of antipsychotics is associated with hip/femur fractures and whether pharmacological differences between antipsychotics are related to the occurrence of fractures. A case-control study was conducted, in which cases were defined as patients with a hip/femur fracture. Each patient was matched to one control patient. The association between use of antipsychotics and the occurrence of hip/femur fractures was evaluated using conditional logistic regression. The study included 44,500 patients from 683 general practices from different geographical areas in the UK, registered within the General Practice Research Database (GPRD). Exposure to antipsychotics was categorized as "no use", "current use" and "prior use". Both current and prior use of antipsychotics were associated with an approximately two-fold increased risk of fractures. After adjustment for possible confounders, a small significant effect remained (Odds Ratios (OR) of 1.3). We did not find an association between dose of antipsychotics, or between the degree of blockade of the alpha-1 adrenoceptor or histamine-1 receptor and risk of fractures. The total number of days of antipsychotic use was significantly associated with an increased risk of hip/femur fractures. We conclude that there is a small increased risk of hip/femur fractures associated with the use of antipsychotics. This risk increases with long-term use.