Sometimes multiple tools may be used simultaneously or in succession in everyday mindreading. Yet surely we rely on somoe tools at some times more than others. The central question I wish to address here, and that I suggest is not well answered, is "Which tool is used when?" More complex and elegant answers to this question await us, but in the meantime I develop several claims in this chapter about which-tool-when contingencies. For instance, I suggest that perceptions of general similarity guide a trade-off between projection (ascribing one's own beliefs and desires to others) and stereotyping. I also claim that displays of self-conscious emotions such as shame, can override inferences of malicious intent from harmful behavior, though only in the short run. My hope, beyond aspiring to be mostly right, is that this discussion will provoke others to address the question of "Which tool when?" To the extent that mind-reading governs much of social life, scholars are well served to understand how perceivers pursue it.