The efficiency of preschool visual screening programmes to detect amblyopia is questionable. In this study such a programme in an inner city was assessed to determine its effectiveness. The results of screening and hospital treatment of 712 patients who were considered to require referral were entered into a database for analysis. Default rates were assessed and the efficacy of treatment determined. The only effective screening test for the detection of amblyopia was visual acuity. A large proportion of referred patients had refractive problems only. High default rates, particularly in geographical areas of lower socioeconomic grading, severely handicapped any attempt to reduce the incidence of amblyopia. A fresh approach to the detection and care of amblyopia in the inner city community is required, perhaps by performing screening of children in their first year of attendance at school to reduce default rates. Cycloplegic refraction of children who are found to have reduced visual acuity before their referral to hospital is also recommended.