BACKGROUND Several methods have been proposed for measuring macular pigment optical density (MPOD). To date, none of the realized techniques can be considered as "gold standard". A key issue for the clinical applicability of a method is its repeatability. In this study, we investigated short- and mid-term repeatability of MPOD measurements using reflectometry. METHODS A total of 12 healthy young subjects were measured 5 times on 5 consecutive days. Repeatability over 6 months was investigated in patients with AMD. The data in AMD patients were taken from a recently published placebo controlled study investigating the effect of lutein supplementation on MPOD (n = 37; [1]). Four measurements over 6 months were used to calculate repeatability. Spectral fundus reflectance of the fovea was measured in a 2.3° detection field with a custom-built fundus reflectometer. Calculation of MPOD was based on a previously published fundus reflectance model. RESULTS The coefficients of variation were 6.2 ± 2.4% and 8.0 ± 5.5% in the healthy and AMD group, respectively. Bland-Altman plots indicate that the difference between measurements at day 1 and day 5 in healthy subjects and day 1 and month 6 in AMD patients was small. The maximum deviation in MPOD in a healthy subject was 0.07 (22.6%), and 0.17 (51.5%) in a patient with AMD. CONCLUSIONS Reflectometry provides adequate short-term and mid-term repeatability for measuring MPOD. Accordingly, the technique makes it possible to monitor MPOD in patients with AMD and to study the influence of supplementation in these subjects. In addition, the variability of the technique is small enough to allow for clinical trials with reasonable sample size.