Since up to 80 percent of pregnant women and 30 percent of neonates may be colonized with genital mycoplasmas, it is difficult to determine whether true infection occurs. The antibody responses to eight serotypes of U. urealyticum were assessed in mothers and infants in 21 cases of neonatal respiratory disease (RD) and 24 normal cases. Among the normal population of mothers and infants, a titer of greater than or equal to 1:32 occurred in 0.25 percent (1/394). In mother-infant paired titers, a fourfold difference occurred in 2.6 percent (5/192). Among 54 RD neonates, 55.6 percent had a titer of greater than or equal to 1:32 compared to only 4.2 percent of normal neonates (p less than .001). Fourfold elevations in antibody titers of greater than 1:32 were observed in the neonate in 52.4 percent of RD cases compared to 0 percent of 24 normal pairs (p less than .001) and in 28.6 percent of mothers of RD neonates compared to 0 percent in normal cases (p = .013). We observed that 43.3 percent of RD neonates with titers greater than or equal to 1:32 died compared to 16.6 percent of RD neonates exhibiting no elevation of antibody response over the maternal level. Among the six who died, 66.7 percent of neonates and 16.7 percent of their mothers had elevated titers, compared to 33.3 percent of 15 surviving infants and 40.0 percent of their mothers. These elevated antibody responses strongly support the concept that U. urealyticum causes infection in the perinatal period in association with neonatal respiratory disease. Since the elevation in titers was detected close to delivery in many cases, the infection may occur in utero.