This review aims to summarise the pathophysiology of C-reactive protein (CRP), the general principles of using it as a prognostic indicator in clinical medicine, and its prognostic role in different diseases. Using the key words ‘C-reactive protein’ with ‘prognosis’, ‘outcome’, ‘response’ or ‘mortality’, clinical studies on uses of CRP as a prognostic test were retrieved from the PubMed database without any language restrictions. CRP is produced by the liver and regulated predominantly by serum interleukin-6 concentrations. It has a half-life of 19 hours and is relatively slow in its onset and offset. Apart from being an inflammatory marker, CRP is increasingly being recognised as a partaker and independent prognostic indicator of some cardiovascular, autoimmune, and neoplastic diseases. A persistently elevated CRP concentration suggests that the underlying inflammatory process is not fully resolved and, in general, this carries a worse prognosis than when CRP responds to the treatment. Because CRP is a non-specific marker of an inflammatory process, it should only be used as a ‘screening’ prognostic indicator to stratify risks for further follow-up or investigation by more specific tests. Furthermore, the limitations of CRP should always be considered when it is used as a prognostic indicator.