This study investigated downscatter effects in cardiac single-photon emission tomographic studies with simultaneous thallium-201/technetium-99m acquisition, and evaluated a previously proposed subtraction technique for downscatter compensation, Ten studies were carried out with different defect sizes and locations and varying activity distributions using four energy windows: 70+/-10% keV, 140+/-10% keV, 100+/-10% KeV, and 103+/-16% keV, The subtraction technique used the 100- or 103-keV data to remove scattered Tc-99m counts from the 70-keV data, The size and contrast of infarcts in the dual-isotope 70-keV image were artificially decreased compared to those in the 140-keV image, caused by scattered Tc-99m counts that were comparable to the primary Tl-201 counts in the 70-keV window. The subtraction technique produced larger defects and more heterogeneous activity in the myocardial wall in dual-isotope 70-keV images compared to the corresponding Tl-201-only images. These artifacts were caused by the markedly different spatial distributions of scattered Tc-99m counts in the 100-keV (or 103-keV) window as compared with the 70-keV window. It is concluded that scattered Tc-99m photons may cause overestimation of ischemia and myocardial viability in simultaneous dual-isotope patient studies. The proposed subtraction technique was inaccurate and produced image artifacts. Adequate downscatter compensation methods must be developed before applying simultaneous Tl-201/Tc-99m acquisition in clinical practice.