Recent studies have shown that accurate characterization of biological samples is only possible by analyzing single cells rather than the average response from thousands of cells, due to the significant heterogeneity in biological samples. Heterogeneity in gene and protein expression at single cell level has been confirmed using different techniques. Therefore, single cell analysis is critical for accurate representation of cell-to-cell variations within a population, which could be masked by average bulk measurements. Single cell analysis can improve data analysis and give some insight for experimental design when dealing with heterogeneous samples. Single cell analysis can be helpful in providing more insight into specific signaling pathways or cellular properties responsible for cell self-renewal capacity or differentiation. Screening sample for this potential rare population of pluripotent cells is critical before their clinical application. Single cell analysis could also improve diagnostics, one example being to distinguish normal and cancer cells at different developmental and metastatic stages. Identifying rare cells such as cancer stem cells is difficult since they only represent a small fraction of the total cell population and unique molecular signatures can be drowned out by noise. Single cell analysis is suggested to enable better identification and targeting of these relatively rare populations in tumors. Studying phenotypes in heterogeneous samples and detection of rare populations requires an information-rich data set of cell characteristics to obtain specificity, which can be aided by multiparameter analysis A longstanding challenge in single-cell analysis is developing specific biomarkers or sets of biomarkers that allow classification of sub- populations of interest, such as cancer stem cells, pluripotent stem cells with high differentiation potential, or immune cells tuned to respond to infections. Sorting is particularly important when nucleic acids are assayed and cells of interest may be rare, and therefore sorting technologies have developed hand-in-hand with analysis approaches. This dissertation reports the development of new tools for label-free multiparameter cell analysis and sorting using its intrinsic biophysical properties.