Here, we report a new handheld endoscopic system for nonlinear multimodal imaging of the head and neck region. It has a long rigid endomicroscopic probe with two versions (0° and 45° bended tip), connecting with a compact scan-head of approx. 10×12×6 cm3 size. The rigid probe is 6 mm in diameter and 24 cm long and allows diffraction-limited multiphoton imaging of tissue with at least 430 μm field of view and sub-micron resolution. The signals of Coherent anti-Stokes Raman Scattering (CARS), second harmonic generation (SHG), and two-photon excited fluorescence (TPEF) are collected by a non-descanned detection path in the scan-head, and the fluorescence of Indocyanine green (ICG) labeled lesions is detected by a confocal descanned configuration. Furthermore, this system is capable of guiding high-power femtosecond laser pulses for tissue ablation without the risk of damaging optical glass components.