Cognitive performance was examined in 131 9-12-year-old children for whom prenatal marihuana and cigarette exposure had been ascertained. The subjects, participants,in an ongoing longitudinal study, were from a low-risk, predominantly middle class sample. The tasks in eluded the WISC-III and a series of tests assessing aspects of cognition subsumed under the rubric of executive function. Consistent with results obtained at earlier ages, discriminant function analysis revealed a dose-dependent association, which remained after controlling for potential confounds (including secondhand smoke), between prenatal cigarette exposure and lower global intelligence scores with the verbal subtests of the WISC maximally discriminating among levels of in utero exposure. In contrast, prenatal marihuana exposure was not associated with global intelligence or the verbal subtests. Rather, this drug was negatively associated with the executive function tasks that require impulse control and visual analysis/hypothesis testing and with a number of WISC subtests requiring the same abilities. The interpretation of these results is discussed in terms of executive function and is related to earlier observations of this sample and to the extant prefrontal and general marihuana literature. (C) 1998 Elsevier Science Inc.