To evaluate if clinically usable estimates of physical performance and level of habitual physical activity are associated with fall risk in elderly men. A population-based sample of 3014 randomly selected men aged 69-80 years was recruited to medical centers in Gothenburg, Malmoe, or Uppsala. The level of physical activity and self-reported falls during the preceding 12 months was evaluated using a questionnaire. The physical performance ability was estimated by measurements of handgrip strength, a timed stands test, a 6-m walking test and a 20-cm narrow walk test. Falls were reported in 16.5% of the men. Fallers performed 6.2+/-19.0% (mean+/-standard deviations; S.D.) less in right handgrip measures, 8.8+/-40.6% slower in the timed stands test, 6.8+/-30.8% slower in the 6-m walking test, and 5.3+/-28.8% slower in the 20-cm narrow walk test (all p<0.001, respectively). The odds ratio for falls among men who performed <-3 S.D. or failed compared to the mean (+1 S.D. to -1 S.D.) in the timed stands test was 3.41 (95% CI 2.31-5.02; p<0.001) and 2.46 (95% CI 1.80-3.34; p<0.001) in 20-cm narrow walk test. There were more fallers that never were physical active (73.0% vs. 65.4%, p<0.001) and who were sitting more (6.4+/-2.5 h/day vs. 6.0+/-2.3 h/day, p<0.05) than among the non-fallers. Fallers scored less than non-fallers in all the estimates of physical performance and they were more sedentary in their life style. The report suggests that clinical usable tests of physical performance and evaluation of habitual physical activity in the clinical situation possibly can be used to predict risk of falls in elderly men.