We have investigated exposure to methylmercury (MeHg) and mercury vapor (Hg0) in pregnant women and their newborns in Stockholm. The women were followed for 15 months post delivery. MeHg, inorganic Hg (I-Hg), and total Hg (T-Hg) in maternal and cord blood were determined by automated alkaline solubilization/reduction and cold vapor atomic fluorescence spectrometry. T-Hg in urine was determined by inductively coupled plasma mass spectrometry. About 72% of the Hg in blood (n = 148) in early pregnancy was MeHg (median 0.94 microg/L, maximum 6.8 microg/L). Blood MeHg decreased during pregnancy, partly due to decreased intake of fish in accordance with recommendations to not eat certain predatory fish during pregnancy. Cord blood MeHg (median 1.4 microg/L, maximum 4.8 microg/L) was almost twice that in maternal blood in late pregnancy and was probably influenced by maternal MeHg exposure earlier and before pregnancy. Blood I-Hg (median 0.37 microg/L, maximum 4.2 microg/L) and urine T-Hg (median 1.6 microg/L, maximum 12 microg/L) in early pregnancy were highly correlated, and both were associated with the number of amalgam fillings. The concentrations decreased during lactation, probably due to excretion in milk. Cord blood I-Hg was correlated with that in maternal blood. The results show the importance of speciation of Hg in blood for evaluation of exposure and health risks.