We have assessed the association between excessive orthostatic changes in blood pressure and risk of incident heart failure (HF) in older, community-dwelling men. This was a prospective cohort study of 3505 men (mean age, 68.5 years), who did not have prevalent HF, myocardial infarction, or stroke. Excessive orthostatic change in blood pressure was defined continuously and categorically as orthostatic hypotension (sitting-to-standing decrease in systolic blood pressure ≥20 mm Hg or diastolic blood pressure ≥10 mm Hg), systolic orthostatic hypertension (increase in systolic blood pressure ≥20 mm Hg, diastolic orthostatic hypertension as diastolic blood pressure ≥10 mm Hg), and orthostatic normotension (neither orthostatic hypotension nor orthostatic hypertension). There was a U-shaped association between orthostatic changes in systolic blood pressure and the risk of incident HF; for diastolic blood pressure, only its fall predicted HF. After adjustment for possible confounders, the hazard ratio (95% CI) for incident HF was 1.65 (1.24–2.18) in men with orthostatic hypotension and 0.90 (0.65–1.24) and 1.88 (1.30–2.73) in men with diastolic and systolic orthostatic hypertension, respectively. Both components of orthostatic hypotension were associated with increased risk, although the systolic component was more predictive than the diastolic component. Both orthostatic hypotension and orthostatic hypertension are associated with risk of incident HF in older men. Our findings suggest that orthostatic hypertension is defined by a rise beyond threshold in systolic blood pressure only. Further prospective studies in diverse cohorts are needed to confirm our findings.