ToC Figure.  A reagent-free hematology analyzer is developed for classifying human leukocyte subpopulations. The system is realized through a cascaded residual neural network and a quantitative phase microscope. Human B and T cells are accurately classified without using chemical labels, outperforming bright-field/dark-field based imaging methods with machine learning. This low-cost and ease-of-use method is suitable for deployment in resource-limited settings.