There is increasing clinical interest in the use of quantitative imaging for radiopharmaceuticals labelled with 111In. Dual-energy-window (DEW) scatter correction is a frequently used component of planar geometric mean quantitative imaging, but it is known that the scatter multiplier k suffers from significant dependence on the characteristics of the scatter medium. Phantom studies with a variety of source geometries were carried out to determine the clinical impact of this dependence on the quantitative accuracy of tumour imaging carried out in conjunction with attenuation correction. Spheres of various sizes (5-20 ml volumes) containing approximately 3.7 MBq (100 microCi) 111In were imaged at a variety of depths (4.8-10.5 cm) within an elliptical water-filled phantom, as well as in air. Geometric mean emission images were acquired using a 20% photopeak window at 247 keV and a 10% scatter window at 205 keV. These emission images were corrected for attenuation using measured 99Tcm transmission data that were scaled to 111In photon energies. Scatter correction was performed in two ways: (1) using the standard DEW method and (2) using a modified DEW method that takes into account benign scatter in the detector crystal. Errors in the activity estimates ranged from -4% to +3% for method 1 in water, and -5% to +3% for method 2 in water. In air, method 1 ranged from -13% to -5%, and method 2 ranged from -10% to -1%. Method 1 was found to yield an accuracy equivalent to that of method 2, except in conditions of very low patient scatter, when the modified method behaved significantly better. We conclude that in a variety of realistic geometries, variations in scatter fraction as determined by the DEW scatter correction method combined with appropriate attenuation correction need not inhibit accurate absolute quantitation of spherical 'tumours' labelled with 111In when using planar imaging.