BACKGROUND: Inappropriate (long-term) use of benzodiazepines (BZDs) is a reason for concern. Several studies have suggested that hospitalisation may be a determinant for initiation of BZD use as well as for long-term use. However, the available evience is conflicting. OBJECTIVE: To determine whether hospitalisation induces initiation of BZD use and subsequent long-term use. METHODS: A retrospective follow-up study was conducted. Randomly, 10,000 patients who had been hospitalised were selected (index date). Non-hospitalised patients, matched on age and gender, were sampled from the same living region and assigned the same index date as the corresponding hospitalised patient. Patients were included if adequate medication data were available from 18 months before to 18 months after the index date. Initiation of BZD use was defined as a prescription for a BZD or BZD-related hypnotic without a prescription for any of these drugs during the prior 6 months. Long-term use was defined as a period of consecutive use for at least 6 months following initiation. RESULTS: In this study, 8,681 hospitalised patients and an equal number of non-hospitalised patients were finally included. Overall, the relative risk for initiation of BZD use was almost twice ashigh [IDR 1.97 (95%CI 1.84-2.10)] among hospitalised patients as in non-hospitalised patients. This relative risk was most clearly elevated during the time window from 3 months before to 3 months after hospitalisation [IDR 4.81 (95%CI 4.08-5.67)]. The relative risk for long-term use during the entire 36-month observation period was not higher [IDR 1.04 (95%CI 0.95-1.13)] among hospitalised patients than among non-hospitalised patients. Within the time window of 3 months before and after hospitalisation,the relative risk for long-term use was significantly lower for the hospitalised group [RR 0.82 (CI 0.69-0.98)]. CONCLUSION: Our results confirm that hospitalisation is associated with an increased risk for initiation of BZD use; the risk is highest during the 3 months just before and after hospitalisation. However, hospitalisation appeared not to be a determinant for long-term use of BZDs.