The continuing high mortality of septic shock has prompted a major effort by the research community to identify novel therapeutic targets. These targets can be conveniently grouped into (1) those derived from microbial components or products; (2) inflammatory mediators; and (3) effector molecules. Many of the experimental, so-called adjunctive agents developed have been monoclonal antibodies or anticytokine molecules of various kinds, and some have progressed into clinical trial. Unfortunately, these trials have failed to show unequivocal survival benefit for patients in shock, prompting a reappraisal of our approach to these agents. In this article, we discuss the possible reasons for these failures: (1) the targets are wrong; (2) the agents are inappropriate; or (3) the trial design is flawed. It would be premature to conclude that adjunctive agents have no future in the therapy of sepsis, but identifying the correct agent, and perhaps more importantly, the correct target population, is going to be more difficult than was at first believed.