Scholars who conduct process tracing often face the problem of missing data. The inability to document key steps in their causal chains makes it difficult to validate theoretical models. In this article, we conceptualize "missingness" as it relates to process tracing, describe different scenarios in which it is pervasive, and present three ways of addressing the problem. First, researchers should contextualize the data generation process. This requires characterizing the process whereby the actors that populate models decide whether to leave traces of their actions and motives. Researchers can thus assess whether or not incentives to produce missingness are compatible with the microfoundations of the theory, and consequently, whether or not missingness is disconfirmatory. Second, researchers may invest in indirect tests of causal mechanisms. Generating out-of-context data about microfoundations offers a plausible window into inaccessible mechanisms. Third, specifying the analytical status of steps in the causal chain allows scholars to make up for deficiencies in evidentiary support.