Extract: With the use of the highly sensitive technique of inactivation by antibody of growth of bacteriophage to which a hapten penicillin is chemically coupled, we found anti-penicillin antibodies (APA) in the sera of all 25 mothers examined and their 27 newborn infants. Precise quantitation of antibodies was not performed in this study, but with the use of optimum dilution of the sera for antibody detection as the indicator, the efficiency, (i.e., a function depending both on amount and on affinity) of antibodies in the sera of the mothers was higher by a factor of 5–10 when compared with the sera of the neonates. The immunoglobulin classes of APA were determined by both the direct method after removal of a single immunoglobulin class by immuno-adsorption on a Sepharose-antiimmunoglobulin column, and by indirect methods such as destruction of heat-labile immunoglobulin (immunoglobulin E is heat labile), and breakdown of macroglobulin (IgM) with 2-mercaptoethanol (2-ME). There was good correlation between the results by both methods, although occasionally some discrepancies were found. Most sera had APA of human immunoglobulins G, M, A, and E (IgG, IgM, IgA, and IgE) classes, but none of the sera of seven mothers and five infants contained IgD APA. More infants than mothers had a high proportion of IgM APA, whereas there were no significant differences between the two groups with regard to IgG, IgA, and IgE APA. In premature infants, all four immunoglobulin classes of APA were found, the youngest infant examined being 6 months gestational age. The proportion of each APA immunoglobulin class in paired mother-infant sera was approximately similar in most cases. However, several pairs were found to have marked discrepancies in the proportion of a specific immunoglobulin class. In three pairs, IgA antibodies were present in mothers' sera diluted 1/500–1/1,000 but not in the infants' sera diluted 1/50–1/100 when other immunoglobulin classes of APA were present. In several pairs, the mother had no IgE APA, whereas it was present in the infant, and, conversely, some mothers had IgE APA, whereas their infants had none. No such remarkable discrepancies were found in the IgG and IgM classes of APA. In the two sets of twins, marked differences were found in each infant pair in the proportion of immunoglobulin class APA. These results indicate that the fetus is able to synthesize IgE antibodies, and are highly suggestive of independent synthesis of IgA, IgM, and IgG antibodies in mother and fetus. These results also suggest that synthesis of the various immunoglobulins by the fetus occurs as early as 6 months gestational age, if not earlier, and that IgE and IgA probably do not pass through the placenta consistently. The presence of antipenicillin antibodies in infants whose mothers had no history of receiving penicillin or its derivatives during pregnancy leads one to speculate that penicillin may be a pollutant of the food we eat and drink. However, the possibility that antibodies cross-reacting with penicillin are present in fetuses must be taken into consideration, particularly as our studies have shown that all neonates have antibodies of various immunoglobulin classes with specificity directed towards the dinitrophenyl hapten.Speculation: Evidence that the immune system of the human fetus is competent and responds to antigenic stimulation in utero indicates that possibilities exist to immunize infants prenatally in order to supply them with adequate and specific active antibodies at birth. The presence of antibodies to penicillin in the sera of all newborn infants studied indicates either that penicillin is a universally present antigen, possibly as a pollutant passing from mother to fetus, or that cross-reacting antibodies to penicillin exist in the fetus.