The technical discourse on nutrition surveillance started decades ago, and the first technical guidelines were proposed in mid-1970s. In spite of this long history, little evidence and consensus exists on the best methods for conducting nutrition surveillance, and on the validity of data produced by these approaches. Multiple nutrition surveillance systems exist in humanitarian settings; however, the validity and usefulness of data produced by these systems are often questionable. In this paper, we outline and define five major methodological approaches to collecting child anthropometric data through surveillance: repeated surveys, community-based sentinel sites, mass screenings, admission data from feeding centers, and data from health clinics. We discuss outstanding methodological and practical challenges with direct implications for quality, validity, and interpretability of collected data and highlight comparative advantages and disadvantages of different methods. We also propose ways forward to building a better evidence base by documenting the strengths and limitations of different approaches, with the eventual goal of achieving consensus on the best ways to collect anthropometric data through surveillance.