Abstract: Jason-3 is equipped with an Advanced Microwave Radiometer (AMR) to remove the signal wet delay caused by precipitable water vapor (PWV). In order to investigate the accuracy of PWV from Jason-3 AMR on a global scale, we adopt PWV observations from 263 radiosonde stations and 103 GNSS stations as reference PWV. These reference PWV are recorded during Jason-3 cycles 0 – 119 and are globally distributed in coastal and island regions. Over 60,000 Jason-3 PWV vs radiosonde PWV comparison points and over 380,000 Jason-3 PWV vs GNSS PWV comparison points are used in this study. For GNSS PWV, four retrieval strategies are used to retrieve GNSS PWV: a combination of two different zenith hydrostatic delay (ZHD) modeling methods (Saastamoinen and ECMWF), and two PWV height reduction methods (Kouba empirical method and ECMWF method to reduce PWV from height of station to sea level). The comparison results indicate that the root mean square error (RMSE) of Jason-3 PWV evaluated using radiosonde PWV is 3.4 kg/m, while the RMSE evaluated with PWV from four different GNSS schemes are in the range of 3.0 – 3.5 kg/m. Specifically, Jason-3 PWV has the best agreement (3.0 kg/mof RMSE) with GNSS PWV derived using Saastamoinen ZHD correction and ECMWF PWV height correction. In addition, the accuracy of Jason-3 PWV increases when the latitude of its footprints or the distance from its footprints to land increases. The correlation coefficient of Jason-3 PWV with radiosonde and GNSS PWV are 0.984 and 0.988, respectively.