Stomach and esophageal cancer are in the top ten most common cancers worldwide, both with high mortality rate. Approximately one-third of these patients have metastases at initial diagnosis and should receive personalized palliative care to improve their remaining life time. However, there is a lack of consensus about personalized palliative care options. This often leads to difficulties in determining the right treatment pathway for individual patients. This study investigates the application of process mining techniques on palliative care pathways for stomach and esophageal cancer to obtain an evidence-based understanding of which palliative treatments are commonly carried out in clinical practice and how they are associated with patients’ survival time. Given the high variability of the treatment pathways, ‘local models’ are derived, rather than end-to-end process models, which are then validated with the aid of physicians. In addition, this study also investigates the use of predictive process monitoring techniques to predict patients’ life expectancy. The results show the benefit of taking the process-flow into account in predicting the outcome of the palliative treatments.