Three different methods of correcting for Anger camera deadtime loss were investigated. These included analytic methods (mathematical modeling), the marker-source method, and a new method based on counting "pileup" events appearing in a pulse-height analyzer window positioned above the photopeak of interest. The studies were done with 99mTc on a Searle Radiographics camera with a measured deadtime of about 6 musec. Analytic methods were found to be unreliable because of unpredictable changes in deadtime with changes in radiation scattering conditions. Both the marker-source method and the pileup-counting method were found to be accurate to within a few percent for true counting rates of up to about 200 K cps, with the pileup-counting method giving better results. This finding applied to sources at depths ranging up to 10 cm of presdwood. The relative merits of the two methods are discussed.