Purpose Burn debridement surgery relies on surgeon experience to remove non-viable tissue. Both excessive and inadequate debridement can have significant consequences, however there are a paucity of tools available to guide intra-operative decision making. Rapid evaporative ionization mass spectrometry (REIMS) links a diathermy electrode to real-time mass spectrometry, providing a continuous molecular profile of tissue being cut. This study investigates if REIMS has the potential to guide burns debridement. Method A single-center, ex-vivo study was performed. Skin samples were obtained from patients undergoing elective and burn debridement surgery. Both normal and burnt tissue samples were analysed using a standardised methodology. Data was visualised with principal component analysis (PCA), and orthogonal projections to latent structures-discriminant analysis (OPLS-DA) was used to interpret variation between groups. Results 11 patients were included, yielding 44 normal tissue data points and 36 burnt tissue data points. Two separate models were generated, one for epidermal and another for dermal depth cuts. Both epidermal and dermal models were significant, with epidermal Q2(cum) = 0.152 and dermal Q2(cum) = 0.147, suggesting promising differentiation between burn and normal tissue. 255 of 900 metabolites were statistically significant at differentiating burnt and non-burnt tissue. For lower m/z features, non-burnt tissue had higher relative abundance; whilst at higher m/z features, burnt tissue had higher relative abundance. In burnt tissue, below 700m/z relative metabolite abundance tends to decrease as burn depth increases. Above 700m/z, metabolite abundance tends to increase as burn depth increases. Conclusion This study demonstrated REIMS has promising ability to differentiate burnt and non-burnt tissue, and identified a number of metabolite trends which contribute to this. It also provides preliminary evidence to support REIMS as a tool to differentiate burn depth.