Body composition and muscle function have important implications for falls and fractures in older adults. We aimed to investigate longitudinal associations between sarcopenic obesity and its components with bone mineral density (BMD) and incident falls and fractures in Australian community-dwelling older men. A total of 1486 men aged ≥70 years from the Concord Health and Ageing in Men Project (CHAMP) study were assessed at baseline (2005–2007), 2-year follow-up (2007–2009; n = 1238), and 5-year follow-up (2010–2013; n = 861). At all three time points, measurements included appendicular lean mass (ALM), body fat percentage and total hip BMD, hand-grip strength, and gait speed. Participants were contacted every 4 months for 6.1 ± 2.1 years to ascertain incident falls and fractures, the latter being confirmed by radiographic reports. Sarcopenic obesity was defined using sarcopenia algorithms of the European Working Group on Sarcopenia (EWGSOP) and the Foundation for the National Institutes of Health (FNIH) and total body fat ≥30% of total mass. Sarcopenic obese men did not have significantly different total hip BMD over 5 years compared with non-sarcopenic non-obese men (p > 0.05). EWGSOP-defined sarcopenic obesity at baseline was associated with significantly higher 2-year fall rates (incidence rate ratio [IRR] 1.66; 95% confidence interval [CI] 1.16–2.37), as were non-sarcopenic obesity (1.30; 1.04–1.62) and sarcopenic non-obesity (1.58; 1.14–2.17), compared with non-sarcopenic non-obese. No association with falls was found for sarcopenic obesity using the FNIH definition (1.01; 0.63–1.60), but after multivariable adjustment, the FNIH-defined non-sarcopenic obese group had a reduced hazard for any 6-year fracture compared with sarcopenic obese men (hazard ratio 0.44; 95% CI 0.23–0.86). In older men, EWGSOP-defined sarcopenic obesity is associated with increased fall rates over 2 years, and FNIH-defined sarcopenic obese men have increased fracture risk over 6 years compared with non-sarcopenic obese men.