Peony is a traditional floral plant in China, whose petals are rich in diverse nutrients and contain high levels of functional components, making it a novel food resource of significant application value. To investigate the characteristics and potential value of peony petals, Ultra-high performance liquid chromatography-tandem mass spectrometry (UHPLC-MS/MS) combined with non-targeted metabolomics was employed to identify the active components in the petals of three peony varieties with distinct flower colours (white Paeonia sufruticosa 'Feng Dan', yellow Paeonia lemoinei 'High Noon', and pink Paeonia sufruticosa 'Ri Xiang').A total of 1,256 metabolites across 16 classes were annotated under both positive and negative ion modes. Significant differences in flavonoids, terpenoids, and alkaloids between groups were identified using univariate t-tests and multivariate analyses. This dataset provides data support for analysing and evaluating the bioactive components in petals of these three peony varieties.