AIM: To investigate cuproptosis-related molecular and immune infiltration in age-related macular degeneration (AMD) development and establish a predictive model. METHODS: The expression profiles of cuproptosis-related genes and immune signature in AMD based on the microarray dataset GSE29801 were analyzed. A total of 142 AMD samples were used to identify the cuproptosis-related differentially expressed genes (Cu-DEGs), together with the immune cell infiltration. To further refine the list of potential genes for AMD diagnosis, three machine learning techniques were used, and an external dataset were applied for confirming the accuracy of the predictive performance. Reverse transcription polymerase chain reaction (RT-PCR) were also performed to examine the level of mRNA of hub genes. The activated immune responses and Cu-DEGs were assessed between AMD and controls. RESULTS: Six genes, including ATP7A, DBT, VEGFA, UBE2D3, CP, SLC31A1, were screened as cuproptosis-signature in AMD via three machine learning methods. Next, SLC31A1 and VEGFA was selected as hub genes by performance evaluation in an external dataset GSE160011, further analysis showed that SLC31A1 and VEGFA were associated with pathways related to immune signaling and immune function, which were then observed in relation to infiltrating immune cells. Finally, the mRNA expression levels of SLC31A1 and VEGFA were significantly higher in laser induced choroidal neovascularization (CNV) group than in control group detected by RT-PCR. CONCLUSION: In this study, the possible relationship between cuproptosis and AMD is expounded systematically. A predictive model is developed to assess the risk of cuproptosis-related genes and their clinical prognostic value in AMD patients.