PURPOSE: To explore the alterations in metabolites of wet age-related macular degeneration (wAMD) by conducting metabolomics in aqueous humor from patients with wAMD and to assess the potential effects of anti-vascular endothelial growth factor (anti-VEGF) on these metabolites. METHODS: Metabolomic analysis was performed on the aqueous humor of 30 patients with wAMD receiving anti-VEGF treatments and 20 controls, via ultra-high performance liquid chromatography tandem mass spectrometry. The aqueous humor samples collected from untreated patients with wAMD were classified as the pre-wAMD group. Accordingly, the samples collected from patients with wAMD receiving one anti-VEGF treatment were designated as the post-wAMD group. Individuals were further classified into responders and nonresponders according to their reaction to the treatment. Principal component analysis, hierarchical cluster analysis, and the Kyoto Encyclopedia of Genes and Genomes annotation and enrichment analysis, were subsequently performed. Machine learning and receiver operating characteristic curve analyses were used to further analyze potential vital metabolites. RESULTS: Among the 1001 metabolites verified in the aqueous humor, 306 compounds separated patients with pre-wAMD from the control group, whereas 68 metabolites differentiated patients with post-wAMD and patients with pre-wAMD. Enrichment in metabolic pathways was noted in ABC transporters, thiamine metabolism, glycerophospholipid metabolism, mammalian target of rapamycin signaling pathway and tyrosine metabolism, and so on. Machine learning and receiver operating characteristic curves analysis suggested that δ-valerolactam could not only distinguish between patients with wAMD and the control group, but also differentiate between patients with post-wAMD and patients with pre-wAMD. Changes in acylcarnitine were observed in anti-VEGF responders with wAMD. CONCLUSIONS: There were noticeable alterations in the aqueous humor of patients with wAMD involving many metabolites that are associated with ABC transporters, glycerophospholipid metabolism, and the mammalian target of rapamycin signaling pathway. It is possible that δ-valerolactam can be applied as a biomarker in wAMD.