Cancer progression is accompanied with alterations in the cell biomechanical phenotype, including changes in cell structure, morphology, and responses to microenvironmental stress. These alterations result in an increased deformability of transformed cells and reduced resistance to mechanical stimuli, enabling motility and invasion. Therefore, single cell biomechanical properties could be served as a powerful label-free biomarker for effective characterization and early detection of single cancer cells. Advances and innovations in microsystems and nanotechnology have facilitated interrogation of the biomechanical properties of single cells to predict their tumorigenicity, metastatic potential, and health state. This dissertation utilized Atomic Force Microscopy (AFM) for the cell biomechanical phenotyping for cancer diagnosis and early detection, efficacy screening of potential chemotherapeutic agents, and also cancer stem-like/tumor initiating cells (CSC/TICs) characterization as the critical topics received intensive attention in the search for effective cancer treatment. Our findings demonstrated the capability of exogenous sphingosine to revert the aberrant biomechanics of aggressive cells and showed a unique, mechanically homogeneous, and extremely soft characteristic of CSC/TICs, suitable for their targeted isolation. To make full use of cell biomechanical cues, this dissertation also considered the application of nonlinear viscoelastic models such as Fractional Zener and Generalized Maxwell models for the naturally complex, heterogeneous, and nonlinear structure of living cells. The emerging need for a high-throughput clinically relevant alternative for evaluating biomechanics of individual cells led us to the development of a microfluidic system. Therefore, a high-throughput, label-free, automated microfluidic chip was developed to investigate the biophysical (biomechanical-bioelectrical) markers of normal and malignant cells. Most importantly, this dissertation also explored the biomechanical response of cells upon a dynamic loading instead of a typical transient stress. Notably, metastatic and non-metastatic cells subjected to a pulsed stress regimen exerted by AFM exhibited distinct biomechanical responses. While non-metastatic cells showed an increase in their resistance against deformation and resulted in strain-stiffening behavior, metastatic cells responded by losing their resistance and yielded slight strain-softening. Ultimately, a second generation microfluidic chip called an iterative mechanical characteristics (iMECH) analyzer consisting of a series of constriction channels for simulating the dynamic stress paradigm was developed which could reproduce the same stiffening/softening trends of non-metastatic and metastatic cells, respectively. Therefore, for the first time, the use of dynamic loading paradigm to evaluate cell biomechanical responses was used as a new signature to predict malignancy or normalcy at a single-cell level with a high (~95%) confidence level. / Ph. D.