It has been suggested that plasma lead (Pb) represents the fraction of circulating Pb that crosses the placenta. Whole-blood Pb levels, the conventional method for measuring circulating Pb, may not adequately reflect plasma Pb levels. The objective of this study was to assess the relationship between whole-blood and plasma Pb in a cohort of pregnant women. A group of 237 pregnant women, recruited in Mexico City from 1997 to 1999, was studied. whole-blood and plasma lead concentrations were evaluated at 12, 24, and 34 wk of gestation by inductively coupled plasma-mass spectrometry. Data were analyzed using a mixed-effects regression model. An exponential relationship was found between plasma and whole-blood Pb levels and significant modification of the association by stage of pregnancy. The association was stronger in the second trimester relative to the first, and it further increased in the third trimester. The model predicts increased plasma Pb levels for a given whole-blood Pb value as pregnancy advances for whole-blood Pb levels greater than approximately 110 microg/L, but not for blood Pb levels less than 100 microg/L. These findings could be due to physiologic changes during pregnancy, such as decreasing hematocrit, saturation of red cell Pb binding capacity, and increased bone resorption or intestinal absorption. Collectively, these data suggest that at elevated maternal blood Pb levels the developing fetus may be at greater risk of Pb exposure from increased maternal plasma Pb than otherwise predicted from whole-blood Pb levels.