The impact of premature birth on associative learning was evaluated using simple delay eyeblink conditioning in which a tone conditional stimulus was paired with an air puff unconditional stimulus. Fourteen preterm (28-31 weeks gestation) and 11 full-term infants completed at least 3 conditioning sessions, 1 week apart, at 5 months of age (corrected age). Preterm and full-term groups demonstrated associative learning, as confirmed by comparison with an unpaired control group. Preterm infants, however, exhibited more variability in their learning rates. The majority of full-term infants and half the preterm infants exhibited rapid acquisition and gradual extinction of conditional responding. A greater proportion of preterm than full-term infants failed to acquire conditional responding within 2 training sessions. Differences in associative learning rates were not the result of differences in arousal or attentional processes. Diversity in acquisition rates exposed an increased risk for disrupted infant learning due to premature birth. (Contains 4 tables and 4 figures.)