A long-standing, but unproven hypothesis is that menopause symptoms cause cognitive difficulties during the menopause transition. This 6-year longitudinal cohort study of 1,903 midlife US women (2000-2006) asked whether symptoms negatively affect cognitive performance during the menopause transition and whether they are responsible for the negative effect of perimenopause on cognitive processing speed. Major exposures were depressive, anxiety, sleep disturbance, and vasomotor symptoms and menopause transition stages. Outcomes were longitudinal performance in 3 domains: processing speed (Symbol Digit Modalities Test (SDMT)), verbal memory (East Boston Memory Test), and working memory (Digit Span Backward). Adjustment for demographics showed that women with concurrent depressive symptoms scored 1 point lower on the SDMT (P < 0.05). On the East Boston Memory Test, the rate of learning among women with anxiety symptoms tested previously was 0.09 smaller per occasion (P = 0.03), 53% of the mean learning rate. The SDMT learning rate was 1.00 point smaller during late perimenopause than during premenopause (P = 0.04); further adjustment for symptoms did not attenuate this negative effect. Depressive and anxiety symptoms had a small, negative effect on processing speed. The authors found that depressive, anxiety, sleep disturbance, and vasomotor symptoms did not account for the transient decrement in SDMT learning observed during late perimenopause.