PURPOSE: To investigate the association between visual impairment and depression and anxiety in older people in Britain. DESIGN: Population-based cross-sectional study. PARTICIPANTS: Thirteen thousand nine hundred people aged 75 years and older in 49 family practices in Britain. METHODS: Vision was measured in 13 900 people aged 75 years and older in 49 family practices taking part in a randomized trial of health screening that included depression (Geriatric Depression Scale [GDS-15]) and anxiety (General Health Questionnaire [GHQ-28]). Cause of visual impairment (binocular acuity less than 6/18) was assessed from medical records. Analysis was by logistic regression (odds ratio [OR] and 95% confidence interval [CI]), taking account of potential health and social confounders. MAIN OUTCOME MEASURES: Levels of depression and anxiety. RESULTS: Visually impaired people had a higher prevalence of depression compared with people with good vision. Of visually impaired older people, 13.5% were depressed (GDS-15 score of 6 or more) compared with 4.6% of people with good vision (age- and gender-adjusted OR, 2.69; 95% CI, 2.03-3.56). Controlling for potential confounding factors, particularly activities of daily living, markedly attenuated the association between visual impairment and depression (OR, 1.26; 95% CI, 0.94-1.70). There was little evidence for any association between visual impairment and anxiety. On the GHQ-28 scale, 9.3% of visually impaired people had 2 or more symptoms of anxiety compared with 7.4% of people with good vision. CONCLUSIONS: Although cause and effect cannot be established in a cross-sectional study, it is plausible that people with visual impairment are more likely to experience problems with functioning, which in turn leads to depression.