Background. There is a lack of data from cohort studies assessing cognitive function prior to and after chemotherapy. We evaluated the effect of self-reported cancer chemotherapy on cognitive function in a cohort assessed at baseline, 4 and 8 years. Methods. Participants were from the population-based PATH Through Life Study. Of the 2,551 participants aged 60-64 at baseline without cognitive impairment, 1,949 completed wave 3 and had data on cancer and chemotherapy and cognitive function. Linear mixed models were used to analyze the data. Results. At wave 3, participants reporting history of chemotherapy (n = 76) had lower scores on memory, processing speed, and executive function compared with those reporting cancer without chemotherapy (n = 289) and no cancer history (n = 1508). After adjustment for depression and disability, effects remained for processing speed and memory. Chemotherapy prior to the study commencement (n = 24), but not between waves 1 and 3 (n = 81), was associated with greater decline in delayed recall (beta = -.21 [95% CI -0.38, -.03], p = .02) and digits backwards alpha = -.05 [95% CI -0.09, -.01], p = .02) over 8 years compared with those with no cancer history (n = 1562). Women reporting chemotherapy for breast cancer after wave 1 (n = 26) had slower choice reaction time (-0.81 (95% CI -1.28, -0.34), p = .001) but did not decline faster on this measure compared with those reporting no breast cancer history (n = 818). Conclusions. Results suggest chemotherapy prior to old age is associated with faster decline in memory in late life but that it does not affect decline in other domains of cognitive function.