PURPOSE: To examine whether delayed rod-mediated dark adaptation (RMDA) in healthy older eyes and those with early age-related macular degeneration (AMD) is associated with AMD incidence and progression at follow-up, as compared with other visual functions reliant on rods, cones, or mixed rod-cone mediation. DESIGN: Longitudinal study over 3 years. PARTICIPANTS: Eyes from adults ≥60 years of age with normal macular health or with early AMD at baseline. METHODS: At baseline, eyes underwent fundus photography to establish AMD presence and severity using the Age-Related Eye Disease Study (AREDS) 9-step classification system. The following visual functions were tested in 1 eye at baseline (the study eye): RMDA, scotopic sensitivity, low-luminance acuity, mesopic contrast sensitivity, mesopic light sensitivity, visual acuity, and contrast sensitivity. Three years later at follow-up, fundus photography and the AREDS classification were repeated. Age-adjusted relative risks and 95% confidence intervals measured the association between visual functions and AMD incidence and progression for the study eye and the fellow eye at follow-up. MAIN OUTCOME MEASURES: Presence and severity of AMD at the 3-year follow-up visit in the study and fellow eyes. RESULTS: Healthy older eyes at baseline with delayed RMDA were 3.54 or 3.40 times more likely to have incident AMD at the 3-year follow-up in the study eye and the fellow eye, respectively, compared with eyes without delays. No other visual functions were associated with AMD incidence, except for low-luminance visual acuity in the study eye. Eyes with early AMD at baseline with delayed RMDA were 3.89 or 2.65 times more likely to have progressed at follow-up for the study and fellow eye, respectively. No other visual functions were associated with AMD progression at follow-up. CONCLUSIONS: Rod-mediated dark adaptation is the only visual function of those tested that was associated with AMD incidence and progression over 3 years in healthy eyes or those with early AMD at baseline. Our results suggest that other visual functions are not useful for understanding AMD incidence and early progression risk. Interventions that eventually are designed to arrest early AMD progression or preventative measures in those at risk should consider RMDA a functional outcome measure. FINANCIAL DISCLOSURE(S): Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.