An endoscope system has an illuminator configured to illuminate an object to be observed with normal light, an image processor that generates an observed image from one field/frame worth of image-pixel signals. The image-pixel signals is read from an image sensor provided in the tip portion of a video scope, and the image processor generates a plurality of spectral images in accordance to the emission of a plurality of narrow-band light. Furthermore, the endoscope system has a detector that detects an operator's aiming spot with respect to at least the plurality of spectral images, a determiner that determines a diagnosis object area that contains the detected aiming spot, and a spectral image processor that extracts a diseased portion with spectral characteristics different from those of the other portions, among the plurality of diagnosis object areas that are defined within the plurality of spectral images.
