Accurate calibration is a critical part in S parameter measurement. The classic TRL method requires many characterization standards to complete the calibration, and assumes the fixtures in the standards are identical. Recently, the 2x-thru calibration has emerged as an attractive alternative because only one THRU standard is needed and that the fixtures in THRU and DUT are allowed to differ slightly. Despite the popularity of 2x-thru calibration in the industry, the underlying algorithms may have long been considered as trade secret, such thatvery few articles in the literature have disclosed how the fixture-mismatch effect can be accounted for. Most of the prior works that are publicly accessible either assume identical fixtures or simply use commercial 2x-thru tools without discussing the calibration algorithm. As such, the accuracy and reliability of 2xthru calibration remain unclarified. In this project, we aim at developing our own 2x-thru calibration algorithms that could handle the problem of fixture mismatch.We propose three levels of algorithms: the 1st level is the one without fixture correction; the 2nd level is one that corrects the S11 mismatch using time-domain techniques, while the S21 and S22 of the fixture are still the same as in THRU; the 3rd level is one that further corrects the S21 and S22. We will use various simulation and measurement test cases to validate the accuracy and significance of the proposed methods. Through this project, we can cultivate engineers with related expertise, and help the industry to solve future problems.