Background: Statin therapy may cause myopathy, but long-term effects on physical function are unclear. Objective: We investigated whether statin use is associated with poorer physical function in two population-based cohorts of older adults. Methods: Data were from 691 men and women (aged 69–102 years in 2005/2006) in the LASA (Longitudinal Aging Study Amsterdam) and 5912 women (aged 79–84 years in 2005) in the ALSWH (Australian Longitudinal Study on Women’s Health). Statin use and dose were sourced from containers (LASA) and administrative databases (ALSWH). Physical function was assessed using performance tests, questionnaires on functional limitations and the SF-12 (LASA) and SF-36 (ALSWH) questionnaires. Cross-sectional (both studies) and 3-year prospective associations (ALSWH) were analysed for different statin dosage using linear and logistic regression. Results: In total, 25 % of participants in LASA and 61 % in ALSWH used statins. In the cross-sectional models in LASA, statin users were less likely to have functional limitations (percentage of subjects with at least 1 limitation 63.9 vs. 64.2; odds ratio [OR] 0.6; 95 % confidence interval [CI] 0.3–0.9) and had better SF-12 physical component scores (mean [adjusted] 47.3 vs. 44.5; beta [B] = 2.8; 95 % CI 1.1–4.5); in ALSWH, statin users had better SF-36 physical component scores (mean [adjusted] 37.4 vs. 36.5; B = 0.9; 95 % CI 0.3–1.5) and physical functioning subscale scores (mean [adjusted] 55.1 vs. 52.6; B = 2.4; 95 % CI 1.1–3.8) than non-users. Similar associations were found for low- and high-dose users and in the prospective models. In contrast, no significant associations were found with performance tests. Conclusions: Two databases from longitudinal population studies in older adults gave comparable results, even though different outcome measures were used. In these two large cohorts, statin use was associated with better self-perceived physical function.