Coning down is commonly used during fluoroscopy to increase image contrast by reducing scatter. However, the resulting image fills only part of a video display whose resolution is limited by line rate and bandwidth. Optical or electron-optical zooming can be used to magnify the collimated image so that it fills a larger fraction of the viewable area of the video frame to make more effective use of the available video-display capacity. Modulation-transfer functions (MTFs) were measured for various zoom factors achieved using a zoom lens and the image-intensifier (II) electronic magnification mode. Significant and continuing improvement in total system MTF was observed up to zoom magnifications of greater than 3.3. For larger zoom factors, the resolution limit becomes dominated by the intrinsic resolving power of the II and by geometric unsharpness rather than by the line rate of the video system. When the MTF at infinite zoom factor, obtained by extrapolation, was divided into the measured MTFs, the resultant MTFz's were shown to scale predictably with zoom factor. Only a slight improvement in MTF was obtained using the II's electronic magnification mode compared to the same magnification using a zoom lens. It is concluded that, if improved image quality is the motivation for the use of coning down in fluoroscopy, then zooming to use fully the available video frame is warranted.