This study assessed changes in balance self-efficacy (BSE) over 1 year in community-dwelling elderly, compared changes in BSE between fallers and nonfallers, and assessed the relationship between specific balance and mobility risk factors for side falls and BSE scores. Elderly fallers (n = 67; 80.2 ± 5.9 years) and nonfallers (n = 75; 79.4 ± 4.9), categorized based on self-reported falls over 1 year, were tested at baseline on postural sway, hip-abduction strength, lateral-stepping velocity, tandem walk, and get-up-and-go and given a BSE questionnaire. Fallers had lower BSE scores than nonfallers did (141.6 ± 33.5 and 154.9 ± 25.4; p = .008). BSE did not change over 1 year. In stepwise regression, BSE scores were predictive of time on the get-up-and-go, mediolateral sway, and tandem walk independent of age, height, and strength (p < .001). The BSE scale might be useful for screening individuals at risk for injurious falls because it is inexpensive and noninvasive.