Primary sensitization to antigens may occur prenatally. We hypothesized that high prenatal exposure to indoor antigens increases the risk for sensitization in newborns in New York City populations with increased risk for asthma. We also investigated whether maternal sensitization is required for in utero sensitization to occur. One hundred sixty-seven pregnant African American or Dominican women residing in northern Manhattan were recruited and antigen was measured from home dust. After delivery, newborn cord and maternal blood were assayed for IgE and mononuclear cell proliferation and cytokine production in response to antigen. Cockroach, mouse, but not dust mite antigens, were commonly elevated in the kitchens and pregnant mothers' beds. Increased mononuclear cell proliferation occurred in 54% of newborns in response to cockroach, 25% in response to dust mite Dermatophagoides pteronyssinus, 40% in response to dust mite D. farinae, and 34% in response to mouse protein extracts. Antigen-induced mononuclear cell proliferation occurred in cord blood even in the absence of antigen-induced mononuclear cell proliferation in the mother. Proliferation in response to antigens did not correlate with IgE levels, but proliferation in response to dust mite extracts correlated with interluekin-5 (IL-5) production in cord blood. These results suggest that (1) high prenatal exposures to cockroach and mouse antigens are prevalent; (2) in utero sensitization to multiple indoor antigens is common, occurs to a different degree than maternal sensitization, and may involve IL-5 upregulation.