Aim To evaluate the effectiveness of a strength and balance enhancing exercise intervention as a means of preventing falls in community‐dwelling elderly Japanese women with a history of falls, while comparing functional fitness, fall and fracture rate in excluded subjects. Methods A 1‐year follow‐up trial was carried out on 105 participants over the age of 70 years, who were randomly assigned to the exercise or education group, and also on 91 women excluded based on the exclusion criteria. The exercise group attended a 60‐min exercise class twice a week for 3 months. Falls, injuries, fractures, and functional fitness assessments were measured at baseline, post‐intervention and 1‐year follow up. Results During the follow up, fall rates were 19.6% in the exercise group, 40.4% in the education group and 40.8% in excluded subjects (χ2 = 7.069, P = 0.029). Compared with the exercise group, the odds ratio (OR) for falls was greater in the education group (OR 2.78, 95% confidence interval (CI) 1.17–6.96) and excluded participants (OR 2.83, 95%CI 1.25–6.80). The OR for fractures was over fourfold greater in excluded participants (OR 4.30, 95% CI 1.02–9.70) than the exercise group. Conclusions The exercise intervention for participants with fall history effectively decreased incidences of falls and fractures. However, fall and fracture rates in excluded people were high. Further research focusing on feasible countermeasures for falls in excluded people who are at high risk of fractures is required. Geriatr Gerontol Int 2013; 14: 285–292.