The effect of lead exposure on neurobehavioral performance is modified by age. Whether educational achievement can serve as an effect modifier on the lead-cognitive performance relationship is examined. The Mini-Mental State Examination (MMSE) and the reading section of the Wide Range Achievement Test-Revised (WRAT-R), a measure of educational achievement, were administered to 256 lead smelter workers. The workers had a mean (standard deviation) age of 41 (7.9) years, education of 10 (2.8) years, employment duration of 17 (8.1) years, current blood lead of 28 (8.8) micrograms/dL, and working lifetime integrated blood lead index (IBL) of 725 (434) micrograms-yr/dL. The median (range) MMSE score was 29 (19 to 30). Multiple linear regression assessed the contribution of age, WRAT-R, education, alcohol intake, cigarette use, IBL, and IBL x WRAT-R on MMSE performance. A significant IBL x WRAT-R interaction examined by stratification found a significant dose-effect relationship between IBL and MMSE, but only in the 78 workers with a WRAT-R reading grade level below 6 years. Workers with higher educational achievement compensated for the effect of lead on cognitive performance.