Falls among the elderly is a major issue in public health, being the cause of debilitating outcomes such as fracture. In this study we investigated whether a common functional variant of skeletal muscle alpha-actinin-3 (ACTN3, R577X) is a risk factor for falls. Homozygosity (XX) for a nonsense polymorphism in the ACTN3 gene (R577X) results in complete ACTN3 deficiency. The XX genotype is present in ~16% Caucasians and has been linked with impaired strength and physical functioning in elderly females. As muscle weakness is an established risk factor for falling, we tested weather ACTN3 R577X genotype is assoicated with falls incidence in older women. Case-control analysis was conducted using two large cohorts of Caucasian postmenopausal women (50+ yrs of age) – the North of Scotland Osteoporosis Study (NOSOS, n=1245) and the Aberdeen Prospective Osteoporosis Screening Study (APOSS, n=2918) – for whom self-reported falls status at multiple time points and DNA samples were available. Cross-sectional analysis of fallers versus non-fallers at baseline and follow-up was performed. Further, individuals who reported having fallen at more than one timepoint (recurrent fallers) were compared with those who did not report falling at any timepoint. Correction for confounders (age, height, weight, alcohol consumption, smoking, socioeconomic status, medications and concomitant medical conditions) was performed by logistic regression. In the APOSS cohort, 22% of individuals reported falling and genotype frequencies were R/R=30%, R/X=49%, XX=21%. In NOSOS, 29% reported falling and genotype frequencies were R/R=29%, RX=52%, XX=19%. For baseline falls, we found association between falling and presence of one or two copies of 577X. Under a dominant model, this result was significant in both NOSOS (OR 1.43, 95% CI 1.00-2.04; p=0.049) and APOSS (OR 1.30, 95% CI 1.04-1.62; p=0.02) independently; meta-analysis provided a pooled OR across both cohorts of 1.33 (95% CI 1.10-1.61; p=0.003). Analysis of independent genotype effects support the dominant model rather than an additive effect, since the pooled OR for comparison of RX versus RR was the same as for XX versus RR (RX: ORpooled = 1.33 [1.09-1.62], p=0.02; XX: ORpooled = 1.34 [1.05-1.71], p=0.005). The effect of the 577X allele on falling was also evident at follow-up falls assessments, though did not reach significance in the individual cohorts likely due to reduced sample size. No significant effect on recurrent falls was observed. These findings indicate that ACTN3 R577X genotype is a risk factor for falling in older women. Carriage of one or two copies of the null variant of ACTN3 increases the risk of falling by as much as 33%. These data support the notion of a genetic component to falls risk and suggests a method for targeted falls intervention programmes in a clinical setting.