Although it has been suggested that IUGR infants, with their more aversive cries and poorer interactive capabilities than normal infants, may be at an increased risk of child abuse, few data are available to either support or refute this hypothesis. Therefore, we conducted a case-control study to determine whether IUGR infants are at an increased risk of being physically abused. Cases were children who: (1) had been born at Yale-New Haven Hospital and were reported because of suspected abuse to the hospital's child abuse committee from Jan. 1974–Dec. 1979, and (2) had evidence of definite or probable physical abuse based on a “blinded” review of the incident by one of the authors. For each case, one control without evidence of abuse was chosen from the hospital's log of births and matched to the case by age, gender, race, method of payment for the hospitalization, and source of pediatric care. To be certain that an association between abuse and IUGR was not missed, 4 different definitions of IUGR were used. Infants were defined as having IUGR if they had either a ponderal index that was below the tenth percentile for gestational age using the Kansas City or Denver growth standards or a birthweight below the tenth percentile using the same two standards. We identified 117 case-control pairs that met the criteria for eligibility. The matched odds ratios for each of the four definitions of IUGR were less than one (ranging from 0.40 to 0.75). Thus IUGR infants occurred more commonly in the controls than the cases and therefore appeared to be at a decreased risk of abuse. The only significant matched odds ratio was for a low ponderal index according to the Kansas City standard (0.4. X2 = .60, p <.05). This result was not affected after controlling for possible confounding factors such as the mother's age or parity, or the child's age at the time of abuse. Based on this carefully designed study, we conclude that IUGR infants are not at an increased risk of physical abuse and may be at a decreased risk.