PURPOSE: The aim of the present study was to determine whether microperimetric parameters could predict the progression of an eye at high risk of age-related macular degeneration (AMD) at 24 months. METHODS: We conducted a multicentric prospective non-comparative open-label study including patients with one eye in stage 4 of the Age-Related Eye Disease Study Group (AREDS) classification, and the other eye in AREDS stage 3 (study eye). A microperimetry examination (MAIA™, CenterVue, Padova, Italy) was performed at baseline and every 6 months during the 2-year follow-up. At the end of the follow-up, each study eye was classified as 'progressive' (i.e. AREDS stage 4) or 'non-progressive' (i.e. AREDS stage 3). RESULTS: A total of 147 patients were analysed, of which 30.6% progressed from AREDS stage 3 to stage 4. The microperimetry criterion 'mean retinal sensitivity' was significantly different at baseline between non-progressive and progressive eyes (p = 0.022), with lower values for the latter. With a threshold for mean retinal sensitivity set at 24.7 dB, diagnostic sensitivity was 80% [95%CI (65.4-90.4)], specificity was 30.4% [95%CI (21.7-40.3)], positive predictive value was 33.6% [95%CI (24.8-43.4)], and negative predictive value was 77.5% [95%CI (61.5-89.2)]. In the multivariate analysis including microperimetric parameters and other routine ophthalmologic examinations, mean retinal sensitivity was the only predictive parameter statistically associated with progression (p = 0.0004). CONCLUSIONS: Our findings are encouraging as regards the use of microperimetry, and mean retinal sensitivity value in particular, to predict the 2-year risk of progression to AREDS stage 4 eye.