Cancer is the second leading cause of death globally, after heart disease. As of 2022, the United States alone reported 1.9 million new cases and 609,360 deaths. Cancer is a deadly and complex disease that is very heterogeneous in nature, making it very difficult to treat. To improve patient outcomes, it is crucial to comprehend how cancer behaves at the molecular and clinical levels. Prognosis, the prediction of disease progression and how a patient will respond to treatment, plays a vital role in designing precision treatments and optimizing care. However, the complexjbity of cancer necessitates the use of sophisticated technologies for data integration and analysis. Computational approaches in the past few years have shown enormous potential in characterizing probable signatures that can predict the cancer prognosis. In this chapter, we have provided an overview of these computational approaches and the various modalities they require for prognosis prediction. These can be broadly classified into four categories (i) data types, including multi-omics, imaging, and clinical data, (ii) traditional statistical and advanced machine learning approaches employed to predict disease progression and survival outcomes, (iii) gene signatures, with a particular focus on immune gene signatures, and (iv) various online resources, packages, and tools developed related to cancer prognosis. Finally, we discussed the various limitations associated with the current computational approaches and the potential future perspectives to address these challenges.