Background Although multiple individual immune parameters have been demonstrated to predict the occurrence of secondary infection after critical illness, significant questions remain with regards to the selection, timing and clinical utility of such immune monitoring tests. Research question As a sub-study of the REALISM study, the REALIST score was developed as a pragmatic approach to help clinicians better identify and stratify patients at high risk for secondary infection, using a simple set of relatively available and technically robust biomarkers. Study design and methods This is a sub-study of a single-centre prospective cohort study of immune profiling in critically ill adults admitted after severe trauma, major surgery or sepsis/septic shock. For the REALIST score, five immune parameters were pre-emptively selected based on their clinical applicability and technical robustness. Predictive power of different parameters and combinations of parameters was assessed. The main outcome of interest was the occurrence of secondary infection within 30 days. Results After excluding statistically redundant and poorly predictive parameters, three parameters remained in the REALIST score: mHLA-DR, percentage of immature (CD10 − CD16 − ) neutrophils and serum IL-10 level. In the cohort of interest ( n = 189), incidence of secondary infection at day 30 increased from 8% for patients with REALIST score of 0 to 46% in patients with a score of 3 abnormal parameters, measured ad D5–7. When adjusted for a priori identified clinical risk factors for secondary infection (SOFA score and invasive mechanical ventilation at D5–7), a higher REALIST score was independently associated with increased risk of secondary infection (42 events (22.2%), adjusted HR 3.22 (1.09–9.50), p = 0.034) and mortality (10 events (5.3%), p = 0.001). Interpretation We derived and presented the REALIST score, a simple and pragmatic stratification strategy which provides clinicians with a clear assessment of the immune status of their patients. This new tool could help optimize care of these individuals and could contribute in designing future trials of immune stimulation strategies.