Children attending non-remedial first and second grades were classified according to the concentration of lead in their shed deciduous teeth. Children in the lowest and highest tenth percentile were studied with a detailed neuropsychological battery under blind conditions. Thirty-nine non-lead covariates were controlled either by matching or in the biostatistical analysis. High lead children tended to have significantly lower IQ scores particularly on the verbal scales of the WISC-R, impaired auditory and language processing, increased reaction times at longer intervals of delay. Their teachers who were blind to the dentine lead levels found an increased incidence of disordered classroom behavior in direct relationship to the concentration of lead in their teeth. Quantitative electroencephalographic analysis demonstrated decreased midline alpha and increased midline delta in high lead subjects. Four years later a subsample of these children was followed up and observed during quiet classroom activity. High lead children tended to spend more time off tasks staring at classmates, out the window or at the observer. These observations demonstrate that lead at doses below those which are associated with frank clinical symptoms produce deficits in intelligence, attention, auditory-language function and disordered classroom behavior.