Aim To identify an appropriate community‐based fall prevention program Methods We introduced two programs to 24 senior centers, “community salons,” in a Japanese city, and carried out a prospective controlled trial between 2004 and 2005. Eight salons (185 participants aged 72.0 ± 7.1 years) received a single‐visit program consisting of one multidisciplinary team visit that included fall risk assessment with feedback and a fall prevention lecture. A total of 16 salons (418 participants aged 73.6 ± 7.4 years) received a year‐round comprehensive program, with visits carried out every 3 months. We compared the fall rates for 1 year between the two programs. Based on the results, we implemented a modified program until 2014 and examined the long‐term consequences. Results In the prospective controlled trial, fall rates did not differ significantly between programs (P = 0.449). Instead, fall rates for both programs decreased significantly by 0.89 (95% CI 0.84–0.94) times each month. Therefore, we implemented a modified version of the single‐visit program. By March 2014, the programs had been delivered to 1863 individuals, and the total number of attendees was 6622. The average attendance frequency per participant was 0.62 times per year. The majority (85.3%) of salons requested the program every year. Overall, the risk of falling (fall rates in the preceding year) decreased significantly as the number of program attendances increased (incident rate ratio = 0.89, 95% CI 0.85–0.92) irrespective of initial program types. Conclusions The programs including fall risk assessment with feedback and a fall prevention lecture reduced falls when embedded into the community, and they were accepted well over the course of 10 years. Geriatr Gerontol Int 2016; ••: ••–••.