The aim of this paper is to analyze merits and demerits of methodological approaches designed for investigations of erythrocyte aggregation - a process, which plays a crucial role in rheological and transport properties of blood. Ideally, erythrocyte aggregation should be characterized in terms of the time-dependent gyration radius of the aggregates and their fractal dimension. Among various experimental techniques suggested so far, only imaging analysis meets this requirement. However, because this technique is designed for investigations of the aggregation process in thin layers of dilute erythrocyte suspensions, aggregation data are affected by cell-wall interactions and, in addition, problems arise when attempts are made to extend these data to whole blood. Interpretation of results obtained by light scattering techniques faces problems associated with effects of multiple scattering, a design of experimental setups and the wavelength on the kinetics of recorded signals. A method based on electric and dielectric properties of blood is advantageous over other methodological approaches, because it provides reliable information about time-dependent and steady-state size and morphology of the aggregates at physiological hematocrits. A common drawback of most methodological approaches is that interpretation of experimental results is based on simplified theoretical models of blood. To avoid complicated physical problems posed by optical, ultrasound, electrical and dielectrical properties of blood, it is suggested to use the adhesion energy as a measure of RBC aggregability.