We studied whether muscle strength measured before a bone fracture predicts mortality following the fracture. The participants were a sub-cohort of 82 people of a total of 493 Evergreen project participants initially aged 75- and 80 years, who had participated in knee extension strength tests as part of the baseline examinations, and who subsequently suffered at least one bone fracture. Maximal isometric knee extension strength was measured at baseline using an adjustable dynamometer chair. Fracture surveillance was carried out from patient records for 5 years, and mortality surveillance from population register for 10 years after baseline examinations. Average time till a fracture occurred after the baseline was 878 days (SD 576). Subsequent to the fracture, 32 deaths occurred. Using gender-specific cut-offs, three equal distribution-based groups were formed. A gradient risk of mortality was found according to baseline strength. The crude mortality rate per 1000 person-months was 15.2 in the lowest 4.9 in the middle and 1.7 in the highest third of baseline knee extension strength. The adjusted relative risk (RR) of death was 4.40 (95% confidence interval, CI 1.40-13.80) in the lowest and 2.39 (95% CI 0.68-8.4) in the middle tertile vs the highest tertile of muscle strength. Poor muscle strength measured before a fracture occurred was a powerful predictor of increased mortality after the fracture. Poor muscle strength may be a good indicator of overall vulnerability and frailty in old age, and strength testing could be helpful in targeting older people for preventive interventions.