Forced-choice preferential-looking (FPL) and pattern-visual-evoked potential (VEP) acuity tests were conducted with 42 children with bilateral moderate to severe visual impairment (age, 4 months to 9 years). Within this group of children, FPL acuity testing was more successful than VEP acuity testing (98% versus 64%). Mean FPL acuity was better than mean VEP acuity (20/155 vs 20/290, P less than .002). For the children who completed both FPL and VEP acuity tests, an acuity difference score (FPL minus VEP) was derived, and compared to difference scores based on normative data. Significant differences were found between visually impaired (VI) and control difference scores (P less than .001). Difference scores of VI were larger than and opposite in sign to the control scores. Rating scores of visual behaviors supported the concurrent validity of the FPL acuities (P = .01), but not the VEP acuities.