One promising approach for removing mercury from coal-fired, utility flue gas involves the direct injection of mercury sorbents. Although this method has been effective at removing mercury in municipal waste incinerators, tests conducted to date on utility coal-fired boilers show that mercury removal is much more difficult in utility flue gas. EPRI is conducting research to investigate mercury removal using sorbents in this application. Bench-scale, pilot-scale, and field tests have been conducted to determine the ability of different sorbents to remove mercury in simulated and actual flue gas streams. This paper focuses on recent bench-scale and field test results evaluating the adsorption characteristics of activated carbon and fly ash and the use of these results to develop a predictive mercury removal model. Field tests with activated carbon show that adsorption characteristics measured in the lab agree reasonably well with characteristics measured in the field. However, more laboratory and field data will be needed to identify other gas phase components which may impact performance. This will allow laboratory tests to better simulate field conditions and provide improved estimates of sorbent performance for specific sites. In addition to activated carbon results, bench-scale and modeling results using fly ash are presented which suggest that certain fly ashes are capable of adsorbing mercury.