Background and purpose: Galectin-3 is a biomarker of atherosclerotic and cardiovascular disease, and may be a useful marker for ischaemic stroke risk. Methods: The Reasons for Geographic and Racial Differences in Stroke (REGARDS) cohort enrolled and examined 30 239 US participants between 2003 and 2007 (41% black, 59% white and 55% in the southeastern stroke belt). Baseline galectin-3 was measured in 526 subjects with incident ischaemic stroke over 5.4 years and in a cohort random sample (CRS) of 947 participants. Cox proportional hazards models were used to calculate hazard ratios (HRs) of ischaemic stroke by quartiles of galectin-3. Results: In the CRS, galectin-3 was significantly higher with older age, black race, female sex, body mass index, hypertension, diabetes mellitus and kidney disease, and also in those who developed incident stroke. Participants with galectin-3 levels in the fourth versus first quartile had a 2.3-fold increased stroke risk [95% confidence interval (CI) 1.6, 3.4] in an unadjusted model. An interaction with age was found (P = 0.06), and therefore age-stratified analyses were performed. Amongst those younger than age 64, baseline galectin-3 in the second–fourth quartiles was associated with increased stroke risk (HR 3.0, 95% CI 1.6, 5.5) compared to the first quartile in an age-, race- and sex-adjusted model. The HR was 2.0 (95% CI 1.0, 4.0) with multivariable adjustment. There was no association amongst older participants. Conclusions: Galectin-3 was associated with incident ischaemic stroke in younger but not older individuals. Confirmation of this finding, and elucidation of its implications for stroke pathophysiology and prevention, is needed.