Within a larger environmental health screening program neurobehavioral measures were taken in 384 6-year-old children (mean age 74 months) in the cities of Leipzig, Gardelegen, and Duisburg. Lead concentrations in venous blood samples (PbB) and urinary mercury excretion in 24-h samples (HgU) were measured as markers of environmental exposure by electrothermal AAS. Dependent variables included two subtests from the WISC [vocabulary (V) and block design (BD)] as well as five tests from the NES2 [pattern comparison, pattern memory, tapping, simple reaction time, and the continuous performance test (CPT; child version)]. In addition, visual functions [visual acuity (TITMUS-test) and contrast sensitivity (FACT)] were tested as covariates. The overall average PbB (geometric mean) was 42.5 microg/l (upper 95% value = 89 microg/l). The overall average mercury excretion (HgU) was 0.16 microg/24 h. Whereas no significant or borderline associations between HgU and any of the target variables was found, significant negative associations were observed between PbB and verbal intelligence (WISC vocabulary but not WISC Block Design) and false-positive responses (false alarms), as well as false-negative responses (miss) in the CPT. Whereas parental education was the most important confounder for WISC performance, visual contrast sensitivity and computer familiarity also proved predictive for performance in several computer-based NES subtests. It is concluded that non-IQ measures, namely measures of sustained attention, are negatively affected in children with 95% of blood-lead levels below 90 microg/l, even after adjustment for intelligence and contrast sensitivity, whereas the causative role of lead in altering IQ functions remains somewhat equivocal, because important covariates could not be controlled for.