Several preclinical models for sepsis have been used in the last decades to successfully unravel the pathophysiologic processes during sepsis. Furthermore, these models for sepsis revealed promising immunomodulating agents for the treatment of sepsis. Nevertheless, several clinical trials evaluating the efficacy of these new anti-inflammatory agents in septic patients showed disappointing results. In this article the advantages and disadvantages of different models for sepsis are discussed. Most models for sepsis lack an infectious focus. Importantly, investigations studying the effects of several immunomodulating strategies have demonstrated strikingly opposite results when using models for sepsis lacking an infectious focus and when using models for sepsis with a more natural route of infection. These differences will be discussed in this article. In general, it is advised to use a combination of models to test a new therapeutic agent, before starting a clinical study evaluating this new therapy.