We developed a forward-looking (FL) multimodal endoscopic system that offers color, spectral classified, high-frequency ultrasound (HFUS) B-mode, and integrated backscattering coefficient (IBC) images for tumor detection in situ. Examination of tumor distributions from the surface of the colon to deeper inside is essential for determining a treatment plan of cancer. For example, the submucosal invasion depth of tumors in addition to the tumor distributions on the colon surface is used as an indicator of whether the endoscopic dissection would be operated. Thus, we devised the FL multimodal endoscopic system to offer information on the tumor distribution from the surface to deep tissue with high accuracy. This system was evaluated with bilayer gelatin phantoms which have different properties at each layer of the phantom in a lateral direction. After evaluating the system with phantoms, it was employed to characterize forty human colon tissues excised from cancer patients. The proposed system could allow us to obtain highly resolved chemical, anatomical, and macro-molecular information on excised colon tissues including tumors, thus enhancing the detection of tumor distributions from the surface to deep tissue. These results suggest that the FL multimodal endoscopic system could be an innovative screening instrument for quantitative tumor characterization.