In the last decade many sources of data other than probability samples have become available as a consequence of the ubiquity of electronic data collection. For example, some vendors and survey organizations have formed large panels of persons who are willing to participate in surveys via the Internet. Many of these sources, despite being large, are not probability samples, but analysts want to project them to full finite populations. This chapter reviews the types of nonprobability data sources that are available and criteria that can be used to judge their quality. We also cover the methods used to make inferences to finite populations using these datasets: quasirandomization, model-based, and doubly robust which is a combination of quasirandomization and model-based techniques.