Microfluidic devices exploit combined physical, chemical and biological phenomena that could be unique in the sub-millimeter dimensions. The current goal of development of Point-of-Care (POC) medical devices is to extract the biomedical information from the blood. We examined the characteristics of blood flow in autonomous microfluidic devices with the aim to realize sensitive detection of interactions between particulate elements of the blood and the appropriately modified surfaces of the system. As a model experiment we demonstrated the fast analysis of the AB0 blood group system. We observed that the accumulation of red blood cells immobilized on the capillary wall leads to increased lateral movement of the flowing cells, resulting in the overall selective deceleration of the red blood cell flow column compared to the plasma fraction. We showed that by monitoring the flow rate characteristics in capillaries coated with blood type reagents it is possible to identify red blood cell types. Analysis of hydrodynamic effects governing blood flow by Finite Element Method based modelling supported our observations. Our proof-of-concept results point to a novel direction in blood analysis in autonomous microfluidic systems and also provide the basis for the construction of a simple quantitative device for blood group determination.