Cervical cancer is the fourth most common cancer in women worldwide and early detection of its pre-cancer lesions can decrease the mortality. Cytopathology, HPV testing, and histopathology are the most commonly used tools in clinical practice. However, these methods suffer from many limitations such as subjectivity, cost, and time. Therefore, there is an unmet clinical need for developing new non-invasive methods for the early detection of cervical cancer. Here, a novel non-invasive, fast, and label-free approach with high accuracy is presented using liquid-based cytology, Pap-smears. CARS and SHG/TPF imaging at one wavenumber of the Pap-smears from patients with specimen negative for intraepithelial lesion (NILM), low-grade (LSIL) and high-grade (HSIL) squamous intraepithelial lesions was acquired. The normal, LSIL, and HSIL cells were selected based on the ratio of the nucleus to the cytoplasm and cell morphology. Raman spectral imaging of single cells from the same smears was also performed providing integral biochemical information of cells. Deep convolutional neural networks (DCNNs) were trained independently with CARS, SHG/TPF, and Raman images, taking into account both morpho-textural and spectral information. DCNNs based on CARS, SHG/TPF, or Raman images have discriminated between the normal and cancerous Pap-smears with 100% accuracy. These results demonstrate that CARS/SHG/TPF microscopy has a prospective use as a label-free imaging technique for fast screening of a large number of cells in cytopathological samples.