Prognostic modelling is used to help identify patient-related factors, and estimate future outcomes, associated with the course of a disease or health condition. They can provide predictions that support decision-making about clinical care of patients. A prognostic analysis thus begins with a clinical question. There are different types of data that can then be used to help answer the question: data collected through an experimental study (e.g. clinical trial), or those derived from an observational study (e.g. disease registry or other routinely available data). A good overview of these two types of data is given in Bosdriesz et al. [1]. Once the clinical question and type of data have been clarified, there are other important considerations to be made at the planning stage before any data analysis is conducted. In this paper, we outline these planning considerations and go on to discuss the modelling and interpretation stages of a prognostic model exercise. The intention is to give an overview of the full process rather than to spend time on one particular stage. We hope that the reader is enticed to further their knowledge by following the references provided throughout the main text as well as those in the Supplementary Information.