In December 2000, the U.S. Environmental Protection Agency (EPA) published a notice of its determination that regulation of coal-fired utilities for mercury is appropriate and necessary as part of the hazardous air pollutant emission regulation for electric utility steam-generating units. To aid in the determination of mercury emissions from these sources, on-line mercury semicontinuous emission monitors (Hg SCEMs) have been developed and tested in recent years. Although Hg SCEMs have shown promise during these previous tests, rigorous field or long-term testing has not been done. In the past year, commercially available and prototype Hg SCEMs have been used by the Energy and Environmental Research Center (EERC) and others at several power plants. As part of the EERC work, Hg SCEMs were operated at a range of conditions and locations. In addition, the Hg SCEMs were operated for up to 1 month. The use of Hg SCEMs at these plants allowed for near-real-time data to be collected under changing plant conditions, as well as during normal ranges of operating conditions. Mercury emission data were obtained from different plants with different configurations. The plant configurations incorporated various pollution control technologies, including selective catalytic reduction (SCR), selective noncatalytic reduction, ammonium sulfate injection for flue gas conditioning, and flue gas desulfurization (FGD). The particulate control devices included electrostatic precipitators (ESPs), a fabric filter (FF), and a venturi scrubber. The testing at these sites included the operation of Hg SCEMs before and after particulate control devices, in wet and dry stack conditions, and at high temperatures (343 °C). The results from these field measurements have provided data that have been evaluated to determine the reliability, variability, biases, and overall capability of Hg SCEMs for monitoring mercury at coal-fired utilities. Even under the best conditions, operation of Hg SCEMs is by no means simple. Their operation at high-dust-condition locations (i.e., prior to the particulate control device), wet stack conditions, and high temperatures has tested the ruggedness of the Hg SCEMs. As a result of this testing, some of the critical factors affecting accuracy, precision, and reliability have been made apparent. The overall capability of Hg SCEMs to produce near-real-time data (data every 5 min) was also evaluated under these conditions. The duration of the tests ranged from short-term periods (1–2 h) up to 1 month. Evaluation of the data over the longer-term tests shows the range of mercury emissions due to the variability of mercury in the coal and the normal range of operating parameters for the unit. The range of data is important when control strategies and managing emission inventories are considered. EPA has not determined what type of monitoring or testing requirements will be included in the upcoming electric utility mercury regulations. There is, of course, interest in requiring the use of Hg SCEMs. However, our research to date has shown that commercially available Hg SCEMs are not capable of operating unattended to produce reliable and accurate emission data.