The effect of maternal exposure to polychlorinated biphenyls (PCBs) on fetal birth weight and gestational age was investigated among humans. Personnel records from PCB manufacturing facilities engaged in producing capacitors using Aroclor-1254 (11097691), Aroclor-1242 (53469219), and Aroclor-1016 (12674112) were examined. Number of pregnancies, birth weight, maternal age, parity, year of birth, race, sex, date of last menses, and parental education data was obtained. There was a total of 388 pregnancies to 354 females and 51 births to 39 females who worked directly in the facility where PCBs were manufactured. These subjects had worked for a minimum of 1 year prior to the birth of the infant and were identified as a high exposure group. There were 337 births to 280 females who worked at locations within the facility where PCBs were not directly used; these were identified as a low exposure group. Birth weights were adjusted for gestational age. The two exposure groups were compared with matched comparisons from the general population. A 53 gram difference was seen between birth weight means of the low and high exposure groups. On the average, the highly exposed workers were older, had more parity, were less educated, and gave birth earlier in the study than the low exposure group. High exposure to PCBs was associated with reduced birth weight even after adjustment for year of birth, maternal parity, and sex of infant. This group also showed shortened gestational age after adjustment for these same variables. Mean gestational age was reduced by 66 days compared to low exposure groups. Average birth weight of low exposure group infants was 66 grams greater than their matched general population comparisons while that for the high exposure infants was 95 grams less than their comparisons. The authors conclude that the mean birth weight difference between the low and high exposure group is the result of a shortened gestational period than a retardation of intrauterine growth.