OBJECTIVE To compare 25-hydroxyvitamin D (25OHD) levels in patients with neovascular age-related macular degeneration (NVAMD) with patients with nonneovascular age-related macular degeneration and control patients. METHODS Medical records of all patients diagnosed with age-related macular degeneration and tested for serum 25OHD level at a single medical center were reviewed. Control patients were selected from patients diagnosed with pseudophakia but without age-related macular degeneration. The lowest 25OHD level available for each patient was recorded. RESULTS Two hundred sixteen patients with nonneovascular age-related macular degeneration, 146 with NVAMD, and 100 non-age-related macular degeneration control patients were included. The levels of 25OHD (mean ± SD) were significantly lower in NVAMD patients (26.1 ± 14.4 ng/mL) versus nonneovascular age-related macular degeneration (31.5 ± 18.2 ng/mL, P = 0.003) and control (29.4 ± 10.1 ng/mL, P = 0.049) patients. The prevalence of vitamin D insufficiency (<30 ng/mL 25OHD), deficiency (<20 ng/mL), and severe deficiency (<10 ng/mL) were highest in the NVAMD group. The highest quintile of 25OHD was associated with a 0.35 (95% confidence interval, 0.18-0.68) odds ratio for NVAMD. CONCLUSIONS This is the largest study to compare 25OHD levels in patients with the different clinical forms of age-related macular degeneration. Mean 25OHD levels were lower and vitamin D deficiency was more prevalent in NVAMD patients. These associations suggest that further research is necessary regarding vitamin D deficiency as a potentially modifiable risk factor for the development of NVAMD.