Vitamin D may be implicated in the pathophysiology of several ocular diseases, but its role in age-related macular degeneration (AMD) remains uncertain. We sought to review systematically the existing evidence to evaluate the association between serum 25-hydroxyvitamin D 25(OH)D levels and AMD. A four-database search (PubMed, ISI Web of Science, Cochrane, and Scopus) was performed from inception to May 2020 using the MeSH terms: ("Macular Degeneration" OR "Age-related macular degeneration" OR "Retinal degeneration" OR "Macula lutea") AND ("Vitamin D" OR "Ergocalciferols" OR "Cholecalciferol" OR "25-Hydroxyvitamin D"). Random-effects meta-analyses were performed to compute 1) the standard mean difference in 25(OH)D concentration between AMD and non-AMD patients and 2) the AMD risk according to serum 25(OH)D levels. Eighteen observational studies enrolling 75,294 patients after a selection process among 375 original abstracts were selected. No significant differences were found, but there appears to exist a trend for late AMD among subjects with a serum 25(OH)D level below 50 nmol/L (odds ratio, 1.8; 95% confidence interval: 1.00-3.24, P = 0.05). There is no clear evidence of a definitive association between serum 25(OH)D and AMD risk, mainly due to heterogeneity in study procedures and lack of longitudinal designs.