492 five to nine year old children participated in a study of the ability of three tests of linear acuity --the Sonksen-Silver Acuity System (SSAS), the Sheridan Gardiner 7-letter chart and the Keelers MK2 chart to identify children requiring specialist visual examination. The ability of the three tests to detect children with visual defects at 6m and of the SSAS in several test test conditions (3m and 6m, monocular and binocular) was investigated at two acuity levels for referral (3/4.5 or 6/9 and 3/6 or 6/12). Although specificity was very high for all tests and test conditions using a 3/6 or 6/12 cut-off for referral, the sensitivity and yield were disappointingly low; higher levels of the latter were achieved at the expense of a degree of specificity using a 3/4.5 or 6/9 cut off. Similarly, sensitivity and yield were greater for the monocular than binocular condition and for the 6m than 3m condition. A test of near vision (SSAS) did not detect any children not already identified by a distance measure. The optometric and developmental rationale behind the differences is discussed.