The Current Employment Statistics (CES) survey collects data each month on employment, hours and earnings from approximately 375,000 business establishments. The CES survey is currently undergoing a sample redesign that includes developing new methods to check the quality of respondent data. In addition to the basic checks for consistency and logic, the new editing tests are designed to identify a variety of plausible patterns in reported data. These tests are longitudinal in nature, meaning that an establishment's current reported values are compared to earlier data for the same establishment. The reported data values are edited using several tests. A change from the current approach used in the survey is that the new edit tests are sequential, passing just one of the tests results in the acceptance of the data. In the current system failing one test results in rejection of the individual microdata. For employment, the edit tests are designed to identify reporting units with a significant difference between the current reported value and earlier sampled employment level. A mismatch or error in this component can result in serious estimation errors. All other data elements are compared to the historical reported values. This paper presents the editing methods being tested and the results of their performance.