This study evaluated the relationships between maternal smoking during pregnancy and 10-year-old children's performance on measures of learning, memory, and problem-solving. In this prospective cohort study, mothers were recruited from an urban prenatal clinic in 1982 and 1983 and observed from their fourth prenatal month until the time of the study. At the 10-year visit, 593 children and mothers were evaluated. The prevalence of tobacco use was high in this cohort: 54.3%, 53.3%, and 60% of the women smoked in the first trimester, third trimester, and 10-year assessment, respectively. After controlling statistically for other prenatal substance use, current tobacco, other substance use variables, and multiple sociodemographic covariates, prenatal tobacco exposure was significantly associated with deficits in learning and memory. Specifically, prenatal tobacco exposure was associated with deficits in verbal learning and design memory, as well as slowed responding on a test of eye-hand coordination. In addition, these children demonstrated a reduced ability for flexible problem solving and more impulsivity, as indicated by an increase in perseverative responses on a card-sorting test. Prenatally exposed children did not show attention deficits or increased activity on a continuous performance test.