The purposes of this study were to determine the umbilical cord blood lead levels in a nonindustrialized city and the effect of place of residence on the blood lead levels of newborns. During spring 1990, we conducted a survey of umbilical cord blood lead levels from 823 live newborns in two hospitals from the Québec City area. The geometric mean of cord blood lead levels was 0.094 mumol/l (95% confidence interval [95% CI] = 0.088-0.099). Less than 1% (95% CI = 0.2-1.7) of the babies had cord blood lead levels at 0.48 mumol/l or greater. We estimate that each year in the Québec City area between 150 and 200 newborns are at risk for developing psychoneurological problems during their first years of life. Frequency distribution of cord blood lead level did not differ significantly among urban, suburban, and rural areas; however, a higher proportion of cord blood lead levels greater than 0.25 mumol/l was observed in the urban area (9.4%), compared with suburban (2.6%) and rural (3.1%) areas. The proximity of potential environmental lead sources (highway, industry, city center) located within 1/2 mile (.8 km) of the mother's residence was associated with a high cord blood lead level. Cord blood lead levels were also elevated for newborns whose mothers lived in newer (< or = 5 y) and older buildings (> or = 50 y). We concluded that the environmental characteristics of the mother's residence are a better indicator of environmental lead exposure than the place of residence.(ABSTRACT TRUNCATED AT 250 WORDS)