Metastasis is a crucial step of malignant progression and remains the primary cause of death for patients with endometrial cancer (EC). However, clinicians presently face the challenge that conventional surgical-pathological variables, such as tumour size, depth of stromal invasion, histological grade, FIGO stage, lymphovascular space invasion or radiological imaging are unable to predict metastatic potential of the primary tumour with accuracy. In the current study, we have compared differential protein expression using primary tumour samples of EC patients diagnosed with (n=24) and without (n=34) lymph node metastasis (LNM). Using matrix assisted laser desorption and ionisation imaging mass spectrometry (MALDI-MSI), we have identified protein signatures from primary tumours to accurately predict LNM. By applying canonical correlation analysis (CCA) based variable ranking approach; we have generated a list of m/z values that could predict the status of LNM in EC. Protein identifications were achieved by in situ MALDI MS/MS in combination with LC-MS/MS. SCiLS lab software was used to visualize the spatial distribution of potential molecular discriminators and their AUC values. Using this approach, we could reliably identify two proteins, which can discriminate EC with and without LNM with an AUC score of 0.2-0.3. Further, the differential expression of proteins was validated by immunohistochemistry. In summary, MALDI-MSI has the potential to identify discriminators of metastasis using primary tumour samples.