The Tekran mercury speciation system is the only commercially viable instrumentation for the routine measurement of mercury compounds in the atmosphere, and is widely deployed in mercury monitoring networks worldwide. To date, however, many key performance measures of the instrumentation have yet to be adequately addressed or documented. While a number of controlled experiments have been conducted in laboratory settings, issues of potential measurement artifacts, non-quantitative collection efficiencies of GOM species, humidity effects, etc. remain to be explored under field conditions. This presentation will address some issues surrounding the accuracy, reproducibility, and robustness of speciated mercury measurements made with the Tekran analytical instrumentation deployed at three AMNet sites operated by NOAA's Air Resources Laboratory: a coastal location (Grand Bay NERR, MS); an inland site in the Mid-Atlantic region (Beltsville, MD); and a high elevation site in the remote free troposphere (Mauna Loa Observatory, HI).