The associations between cigarette smoking history and later cognitive performance were examined among 3,429 Japanese-American participants of the Honolulu Heart Program (HHP) and its extension, the Honolulu-Asia Aging Study (HAAS). Cognitive performance was measured by the Cognitive Abilities Screening Instrument (CASI), administered as part of HAAS (mean age at HAAS exam (standard deviation (SD)): 77.7 (4.6) years). Information on smoking history was collected during the first and third HHP exams (mean age (SD) at Exam III: 58.6 (4.7) years). Compared with never-smokers, those who had smoked continuously between Exams I-III and those who had quit smoking during that period had significantly lower CASI scores, after adjustment for age, education, Japanese acculturation, and Exam III alcohol intake. In multiple logistic regression controlling for the above covariates, a significantly higher risk of cognitive impairment (CASI score < 82) was associated with continuous smoking (odds ratio (OR) = 1.36, 95% confidence interval (CI) 1.10-1.69) and quitting between Exams I-III (OR = 1.36, 95% CI 1.03-1.80) compared with never smoking. This excess risk of cognitive impairment among continuous smokers and Exam I-III quitters was slightly diminished by further adjustment for body mass index and several vascular covariates. Additional analyses suggested a reduced risk of cognitive impairment among the longer-term quitters. This study suggests a positive association between smoking during middle age and later risk of cognitive impairment.