We determined the secular trend in blood lead levels in a cohort of 321 children born in Mexico City between 1987 and 1992. Blood lead level was measured every 6 months during a 10-year period. We modeled the effect of yearly air lead concentration nested within the calendar year in which the child was born, family use of lead-glazed pottery, socioeconomic status, year in which the child was born, age of the child at the time of blood lead measurement, place of residence, and an indicator variable for subjects with complete or incomplete blood lead values. The yearly mean of air lead of the Valley of Mexico decreased from its highest level of 2.80 microg/m3 in 1987 to 0.07 microg/m3 in 2002. The contribution of air lead to blood lead according to year of birth was strongest for subjects born in 1987 and fell to nearly zero for children born in 1992. The geometric mean of the entire cohort rose from 8.4 microg/dL in the first year of life to 10.1 microg/dL in the second and decreased thereafter until it reached 6.4 microg/dL at 10 years of age. Children of families who used lead-glazed ceramics had blood lead levels 18.5% higher than did children of nonusing families. Children who belonged to the lowest socioeconomic levels had blood lead levels 32.2% higher than did those of highest socioeconomic levels. Children who lived in the northeast part of the city had blood lead levels 10.9% higher compared with those who lived in the southwest.