We demonstrate a novel label-free OCT method allowing optical detection and differentiation of moving micro-objects, such as blood cells. In this study we use phase-sensitive Fd-OCT/OCM system with broadband light source (axial resolution: 3 μm in tissue and lateral resolution 4-8 μm) and fluidic sample setup. The novel part of this method is that the optical identification is based on optical signal coming from optically uniform scattering media localized beneath the flowing/moving objects and not from the objects itself. This signal reveals as an enhancement in speckle pattern on intensity images and non-zero phase change on phase images (phase-change). Statistical parameters of such signals depend on the features of the object, like its size, shape, internal structure, etc. In order to demonstrate the effectiveness and reliability of proposed method, we performed an experiment to differentiate erythrocytes from leukocytes. Obtained OCT cross-sectional images present signal enhancement in the scattering base, both on intensity and phase-change images. This modulation signals, corresponding to erythrocytes and leukocytes, are significantly different and could be easily distinguished qualitatively. Statistical parameters used for the analysis also represent satisfactory separation to distinguish between different kinds of cells. Above-mentioned results of differentiation are presented in this paper.