Exposure to tobacco smoke is the most important environmental health hazard. Studies of the effects of parental smoking on pregnancy have generally relied on questionnaires to measure exposure. However, parental interviews are susceptible to information bias. The use of biochemical markers of cigarette smoke absorption, such as carbon monoxide (CO), has thus been suggested to be more accurate. The aim of this study was to determine whether exposure to tobacco smoke in late pregnancy may be reliably estimated by measuring CO in the newborn's breath. A portable automated end-tidal CO analyzer that rapidly samples and analyzes newborn's breath at the bedside (Baby's Breath™, Natus Medical, Inc., San Carlos, CA) was used. Forty-three full-term, healthy non-jaundiced newborns and their parents had end-tidal CO measurements[corrected for inhaled air (ETCOc)], performed within 12 hours of birth. All neonates had blood group and direct Coombs' test, as well as serum bilirubin, hematocrit and reticulocyte count, determined in order to rule-out hemolysis which could contribute to ETCOc levels. The mean±SD ETCOc of the mothers who smoked was significantly higher (p<0.01) than non-smokers(7.1±4.9 vs. 1.3±0.8 ppm). The mean±SD ETCOc of the newborns whose mothers smoked was significantly higher (p<0.05) compared to infants of non-smokers (6.8±5.7 vs. 1.5±0.8 ppm). Linear regression analysis for maternal and newborn ETCOc measurements showed a good correlation [Newborn ETCOc = 0.598 + 0.735 Maternal ETCOc (r=0.76)]. Linear regression analysis for paternal and newborn ETCOc measurements (excluding infants whose mothers smoked) showed a poor correlation [Newborn ETCOc = 1.531- 0.013 Paternal ETCOc (r=0.33)]. A similar poor correlation was found for paternal and maternal ETCOc (r=0.39). We conclude that ETCOc measurements in the newborn are well correlated with the maternal smoking. Paternal smoking, a source of exposure to passive smoking, did not result in increased ETCOc levels in the mother or her newborn. Measurements of newborn ETCOc may be used as a noninvasive means to estimate exposure to maternal tobacco smoke immediately before delivery for both patient education and research purposes.