Bailey-Lovie-Ferris visual acuity charts and Teller visual acuity cards were used to compare recognition and grating visual acuity at near testing distances in 32 children with anisometropic amblyopia without strabismus. Appropriate optical corrections were worn. Test-retest intraobserver reliability was higher for letters (r =.95) than for gratings (r = .68). Using 20/30 visual acuity or better as the criterion for normal visual acuity, eight eyes with letter visual acuities ranging from 20/42 to 20/138 would have been inaccurately found to be normal by using the Teller visual acuity cards alone. Grating visual acuity measurements tended to be better than letter visual acuity; and, in general, they did not worsen proportionately with poorer letter visual acuity.