Two years' experience with a preferential looking technique for visual acuity determination was evaluated. Ninety-three patients were tested during this period, but selected groups (34 patients) were reported on earlier. In 59 patients aged 2 months to 13.75 years 131 testing procedures were performed. Diagnostic groups included uncomplicated strabismus (n = 26), strabismus combined with neurological disturbances (n = 16), neurological disturbances alone (n = 11) and other ocular diseases (n = 6). In addition, several of the children were mentally retarded. In 52 of the 59 patients in this study valuable results were obtained, while no useful information could be gathered from the remaining 7 patients. The findings influenced therapy strategies in 24 patients, and subsequent testing reflected therapeutic effectiveness. The test was successful in 7 of 8 patients referred from other clinics for estimation of visual function. Although established acuity thresholds should be regarded as conservative estimates, and not directly comparable to Snellen letter acuity notations, the preferential looking technique proved to be quite valuable in a paediatric ophthalmology unit. Main indications were amblyopia detection, control of therapeutic measures in strabismics, including pre-operative evaluation, and estimation of visual capacity in children with psychomotor retardation.