Flexible micro-endoscopes produce images that are obscured by a `honeycomb' pattern due to the negative space between the individual optical fibers contained in the imaging conduit. This pattern is found to exhibit a spatial frequency that is visibly distinct from that of the imaged object. By applying a frequency filter, it was possible to remove the honeycomb pattern without significant degradation to the visual quality of the image. This process will greatly increase the perceived quality of the information being obtained by the endoscopes and aid in their effective use for medical diagnosis. The technique described employs Fourier spectral analysis to determine the `noise' component in the original image. A discrete band-reject frequency filter was designed by visually examining the spectral information and creating the necessary filter to block out the undesired frequency band. The honeycomb pattern was no longer distinguishable after applying this filter to a test gray-level image of an Air Force calibration target. Next, the filter was applied to the individual color planes of a sample color image. The color planes were recombined to produce a full-color image free from interference. A full description of the methods involved is presented.