Researchers do not have as much control over field period for self-administered surveys as they do for telephone surveys. Progressive increases in the use of Web as a data collection mode for general population surveys have resulted in growing interest in how long a web survey should be in the field and how many contact attempts should be made to ensure representation of various population subgroups. In this presentation, we examine field duration and contact attempts required for reaching various subgroups in a web survey. We examine the key demographics of early and late responders, as well as key survey variable measurements at various points in time. We utilize data from the Pew Research Center’s American Trends Panel (ATP) for this analysis. ATP is a nationally representative probability web-based panel comprising of approximately 7,000 panelists. The field period for each wave generally lasts about 3 weeks. Our analysis is based on aggregated data for the last three waves of data collection for ATP. These analyses will inform survey researchers on the impact of contact protocol and survey field period on representativeness for web surveys.