Several studies suggest that clinical researchers are prone to overestimating the chance that scientific protocols will succeed, say, by confirming a treatment hypothesis or reaching recruitment targets. In this essay, we address the normative question of whether such "unrealistic optimism" is harmful in clinical research. We entertain five plausible defenses of unrealistic optimism: agnosticism (optimism cannot be objectively assessed), skepticism (expressions of unrealistic optimism are difficult to interpret, since researchers are not accustomed to thinking in terms of probability), contrarianism (unrealistic optimism is salutary), denialism (any effects of unrealistic optimism are neutralized by oversight mechanisms), and fatalism (nothing can be done to alter unrealistic optimism or its effects). Though each argument has force, we find each insufficient to dispel moral concerns about community-wide unrealistic optimism in research. We close by describing how each argument might inform the study and moral evaluation of unrealistic optimism in research.