Binocular grating acuity was tested in 138 low birth weight (LBW) neonates (birth weights ranging from 1500 to 2500 g) by means of the prototype version of the Acuity Card Procedure. No surrounding screen was used. Mean visual acuity of 107 neonates successfully assessed at mean corrected ages of -1.9 weeks (+/- 1.9 weeks) amounted to 0.58 cycles/degree (S.D. 0.71 octaves). Success rate was 77.5%. Mean postnatal age was 2.3 weeks (+/- 1.6 weeks). Acuity values of various subgroups ranged between 0.68 cycles/degree (S.D. 1.3 octaves) in low-risk, small for gestational age (SGA) preterms (n = 7), to 0.56 cycles/degree (S.D. 0.7 octaves) in SGA fullterms (n = 34), independent whether at low-or at high-risk. These differences were not significant, although with multiple regression analysis with adjustment for corrected age of testing, mean acuity of low-risk preterms was slightly better than of low-risk fullterms (P = 0.055). No significant change of acuity over corrected age could be demonstrated, except for a slight progress (r = 0.57; P less than 0.05) in the subgroup of 13 low-risk fullterms. The high variability of acuity values in neonates and the slow acuity development at term age hamper assessment of differences between various subgroups of neonates.