The study was carried out in the municipality of Aarhus, a city of 250,000 inhabitants. The study was designed as a cross-sectional cohort study of school children in first grade in 1982-83. A total of 2,412 children were contacted and asked to submit their shed teeth to the teacher, and 1,291 children delivered at least one usable tooth (response rate, 54 percent). The lead level in circumpulpal dentin showed an average of 10.7 micrograms/g. Eight percent of the children (N = 110) had a lead level above 18.7 micrograms/g and were selected as a "high" lead exposure group. This group was matched by sex and socio-economic status of the parents with control children with a dentin lead level below 5 micrograms/g. Following a detailed interview with the parents, children were excluded from the study if medical risk factors were present. A clinical psychologist, blind to the lead data, administered selected psychometric tests to 162 of the children selected. The high-lead children scored lower on the WISC when compared to low-lead children, especially on the Verbal IQ (p less than 0.001) and Full Scale IQ (p less than 0.01). No significant difference was seen between the high- and low-exposure groups on the Performance IQ and on several experimental tests. Impaired function associated with lead exposure was also found on the Bender Visual Motor Gestalt Test (p less than 0.001) and on a behavioral rating scale (p less than 0.01). These results remained statistically significant even after controlling for socio-economic status and other confounding variables.