BACKGROUND: Reduced quadriceps strength is an early finding in subjects with knee osteoarthritis, but it is not clear whether it is a cause or a consequence of knee osteoarthritis. OBJECTIVE: To determine whether reduced functional performance in the lower extremity predicts the incidence or progression of radiographic knee osteoarthritis. DESIGN: Prospective, epidemiological, population based cohort study. PATIENTS: 148 subjects (62 women), aged 35-54 (mean 44.8), with chronic knee pain from a population based cohort. MEASUREMENTS: Predictors analysed were age, sex, body mass index, baseline knee pain, and three tests of lower extremity functional performance: maximum number of one-leg rises from sitting, time spent walking 300 m, and timed standing on one leg. Weightbearing tibiofemoral knee radiographs were obtained at baseline and after 5 years (median 5.1, range 4.2-6.1), and classified according to Kellgren and Lawrence as no osteoarthritis (Kellgren and Lawrence = 0, n = 94) or prevalent osteoarthritis (Kellgren and Lawrence >/=1, n = 54). RESULTS: Fewer one-leg rises (median 17 v 25) predicted incident radiographic osteoarthritis five years later (OR 2.6, 95% CI 1.1 to 6.0). The association remained significant after controlling for age, sex, body mass index, and pain. No significant predictor of radiographic progression in the group with prevalent osteoarthritis was found. CONCLUSION: Reduced functional performance in the lower extremity predicted development of radiographic knee osteoarthritis 5 years later among people aged 35-55 with chronic knee pain and normal radiographs at baseline. These findings suggest that a test of one-leg rises may be useful, and interventions aimed at improving functional performance may be protective against development of knee osteoarthritis.