Aim. To assess the projected needs for cataract surgery by lens opacity, visual acuity, and patient concern. Methods. Data were collected as part of the Melbourne Visual Impairment Project, a population based study of age related eye disease in a representative sample of Melbourne residents aged 40 and over. Participants were recruited by a household census and invited to attend a local screening centre. At the study sites, the following data were collected: presenting and best corrected visual acuity, visual fields, intraocular pressure, satisfaction with current vision, personal health history and habits, and a standardised eye examination and photography of the lens and fundus. Lens photographs were graded twice and adjudicated to document lens opacities. Cataract was defined as nuclear greater than or equal to standard 2, 4/16 or greater cortical opacity, or any posterior subcapsular opacities. Results. 3271 (83% response) people living in their own homes were examined. The participants ranged in age from 40 to 98 years and 1511 (46.2%) were men. Previous cataract surgery had been performed in 107 (3.4%) of the participants. The overall prevalence of any type of cataract that had not been surgically corrected was 18%. If the presence of cataract as defined was considered the sole criterion for cataract surgery with no reference to visual acuity, there would be 309 cataract operations per 1000 people aged 40 and over (96 eyes of people who were not satisfied with their vision, 210 eyes of people who were satisfied with their vision, and three previous cataract operations). At a visual acuity criterion of less than 6/12 (the vision required to legally drive a car), 48 cataract operations per 1000 would occur and people would be twice as likely to report dissatisfaction with their vision. Conclusions. Estimates of the need for cataract surgery vary dramatically by level of lens opacity, visual acuity, and patient concern. These data should be useful for the planning of health services.