OBJECTIVE: To investigate whether there is a difference in the risk of progressing to BPH-related prostatic surgery between patients using alpha-blockers and patients using the 5-alpha-reductase inhibitors (5-ARIs). METHODS: A population-based cohot study was conducted, using data from the PHARMO Record Linkage System. We identified 5671 patients (> or =50 years old, no history of using both alpha-blockers and 5-ARIs, more than one year of database history prior to the first date of BPH drug-dispensing), who filled at least one prescription for either alpha-blockers (alfuzosin, tamsulosin, terazosin) or 5-ARIs (finasteride). The incidence of BPH-related surgery was compared between patients treated with alpha-blockers and patients treated with 5-ARIs. RESULTS: The cumulative incidence of BPH-related prostatic surgery was 15.2% and mainly involved transurethral resection of the prostate (TURP) (13.4%). Patients using alpha-blockers had a significantly increased risk of BPH-related prostatic surgery compared to patients using 5-ARIs, which remained after adjusting for age, calendar time, type of prescriber and chronic disease score (adjusted HR: 1.52, 95% CI: 1.24-1.88). The difference between alpha-blockers and 5-ARIs was sustained after stratification of time period (<1995, > or =1995) and exclusion of patients with prostatic surgery within one month of treatment initiation. CONCLUSIONS: It is concluded that alpha-blocker treated patients had a higher risk of BPH-related surgery compared to 5-ARI treated patients. Additional research on the long-term outcomes and risk factors for the natural progression of BPH is necessary to identify the optimal medical treatment for BPH patients according to their baseline characteristics.