This randomized, double-blind trial tested the hypothesis that the addition of 0.2% docosahexaenoic acid (DHA, 22:6n-3) from marine oil to commercially-available preterm and term formulas with > or = 3% linolenic acid (18:3n-3) would enhance novelty preference and visual attention of preterm infants. Among preterm infants cared for in our center, study infants were a select group considered to be at lower risk for developmental delay. Study infants received their assigned diet (control, DHA-supplemented) from a mean postnatal age of 25 d until 9 mon past term. At 6.5, 9, and 12 mon past term, they were tested for visual recognition memory (novelty preference) and attention with the Fagan Test of Infant Intelligence. The effects of DHA supplementation were analyzed by repeated measures analysis of variance. In paired comparisons of novel and familiar stimuli, DHA-supplemented and control infants had the same novelty preference, but supplemented infants had more discrete looks to both novel (P < 0.03) and familiar (P < 0.02) stimuli and a shorter overall look duration (P < 0.03). These data are analogous to those from n-3-deficient and n-3-fed monkeys in that the group with better DHA status had shorter overall look duration. Because shorter look duration has been associated with more rapid information processing, preterm infants fed formulas with only linolenic acid may have had slower information processing than those fed DHA.