Although measles was declared eliminated in the United States in 2000, it has reemerged in the United States and abroad. In the first nine months of 2019, a total of 1249 cases of measles were reported in the United Statesdthe highest number on record since 1992 and nearly double the number of cases seen in any year in the past decade. 1,2 In the Democratic Republic of the Congo, a striking 4000 deaths were attributed to measles in 2019, and more than 200,000 measles cases have been reported year to date. 3 Efforts to vaccinate Congolese children and supply medicines are ongoing as health authorities attempt to gain control of the epidemic. 4 In the United States, the resurgence of measles has been attributed to reduced vaccination rates driven by vaccine hesitancy and misinformation. 5,6 Fear of vaccines rose in 1998 in response to the now widely discredited claim that the MMR vaccine causes autism. 7 Readers are referred to my June Note, 8 within which I described ways to engage with those who remain opposed to vaccination to help mitigate their concerns. Last month, infectious diseases experts participated in a global "Twitterstorm" during IDWeek to advocate for vaccination and spread accurate information using the hashtag #WhyIVaccinate. 9 In addition to these efforts, we turn to pharmacoepidemiologists to supply objective, rigorous scientific evidence about vaccine usage and safety, to disseminate this information to the public, and to counter misinformation. Strictly speaking, pharmacoepidemiology refers to the study of the use, effectiveness, and safety of pharmacologic agents (ie, drugs, biologics, vaccines) in large, well-defined populations. Pharmacoepidemiology focuses on groups rather than on individuals. Studying pharmacologic agents under real-world conditions in sufficiently large samples or populations should yield inferences about their effectiveness, tolerability, and safety that are more convincing and reproducible. Whenever one reads a case report or even the results of a well-designed and carefully executed clinical trial, certain questions tend to surface: Is this finding truly representative or generalizable? In a given trial, has randomization adequately equated the limited samples or has it inadvertently introduced random or sampling errors? When groups are compared, were the sample sizes really large enough to remove the impact of any relevant differences? Clearly, the larger the sample size, the less likely it is that sampling or random errors will affect inferences. Currently, many pharmacoepidemiological studies utilize data from electronic health records, registries, and other databases. These sources are usually chosen because the information they contain is available and bounded by defined time periods. Some obvious limitations of data from such sources are that data collection and entry were not intended for research purposes. Most often, available data were entered by different people and usually for administrative or medical billing purposes. Important data may be missing from t...