Background Alpha-1 blockers are the first option for the medical treatment of benign prostatic hyperplasia (BPH), and differ in their likelihood of causing abnormal ejaculation. Alpha 1A-selective blockers may result in reduced or absent seminal emission via inhibition of smooth muscle contraction. A new alpha-1 blocker, silodosin is 162 times more selective for alpha-1A than for alpha-1B, and is approximately 55 times more selective for alpha-1A than for alpha-1D. In contrast, naftopidil is an alpha 1D-selective blocker, which has been recently reported to less likely induce ejaculatory disorders. To investigate clinical efficacies and impacts on sexual function of these alpha-1 blockers in patients with BPH, we performed a multicenter randomized trial. Methods Eighty-nine patients with of lower urinary tract symptoms (LUTS)/BPH who had International Prostate Symptom Scores (IPSS) of 8 or more were randomly assigned to receive silodosin (8mg/day, n = 48) or naftopidil (75mg/day, n = 41). Before and 4, 8, and 12 weeks after treatment, IPSS, and Quality of Life (QOL) were employed to assess LUTS. Also, IIEF-5 and an original questionnaire were used to evaluate erectile function and ejaculation, respectively. Results Efficacies on LUTS of these blockers are almost equivalent, with a small advantage of silodosin on voiding symptoms. A significant impairment of ejaculation (decreases of ejaculatory volume) was observed in the silodosin group (p < 0.01). Orgasm and satisfaction on ejaculation in the silodosin group was significantly inferior to those in the naftopidil group. Conclusions Alpha 1D-selective blockers may possess superior property of preserving sexual function, compared with Alpha 1A-selective blockers.