Since 1939, the Current Employment Statistics survey has used a single page form for capturing data. Over time, problems with the design of the form have become apparent. In 2009, research on how to improve the form was completed. Prototypes based on recent research were created and fielded for testing during 2011 and 2012. The test consisted of systematic random samples which were selected and sent to regional data collection centers each month. Samples were stratified by geography, industry, single location vs. multiple locations, and size of firm. Results show improvements in interview times and reduced item non-response.