Conventionally, liver fibrosis is diagnosed using histopathological techniques. The traditional method is time-consuming in that the specimen preparation procedure requires sample fixation, slicing, and labeling. Our goal is to apply multiphoton microscopy to efficiently image and quantitatively analyze liver fibrosis specimens bypassing steps required in histological preparation. In this work, the combined imaging modality of multiphoton autofluorescence (MAF) and second-harmonic generation (SHG) was used for the qualitative imaging of liver fibrosis of different METAVIR grades under label-free, ex vivo conditions. We found that while MAF is effective in identifying cellular architecture in the liver specimens, it is the spectrally distinct SHG signal that allows the characterization of the extent of fibrosis. We found that qualitative SHG imaging can be used for the effective identification of the associated features of liver fibrosis specimens graded METAVIR 0 to 4. In addition, we attempted to associate quantitative SHG signal to the different METAVIR grades and found that an objective determination of the extent of disease progression can be made. Our approach demonstrates the potential of using multiphoton imaging in rapid classification of ex vivo liver fibrosis in the clinical setting and investigation of liver fibrosis-associated physiopathology in animal models in vivo.