A large number of immunomodulatory therapies has been evaluated in patients with severe sepsis and septic shock. Until recently, none of these treatments has ever demonstrated any benefit in terms of decreased mortality. Many biases could interfere with the results of these clinical trials linked to poor comprehension of immune response, pharmacological errors, selection bias, and mistakes in the evaluation of the patients and in the interpretation of the results. Based on these methodological flaws, the authors try to define directions for future clinical trials.