The present article reports a method for the average grain size evaluation of superconducting nano-particles through their magnetic properties. The use of SQUID magnetometry to determine the average MgB2 particle size was investigated and the results compared with those achieved through other techniques. In particular the data obtained from zero field cooled magnetization measurement as function of the temperature were compared with the results obtained by scanning electron microscopy and Brunauer-Emmett-Teller techniques. The particle magnetization was measured by a commercial SQUID magnetometer in magnetic field (1 mT) and temperatures ranging from 5 to 50 K dispersing the powders in a grease medium. The grain size is obtained by fitting the data taking into account the Ginzburg-Landau temperature dependence of the London penetration depth. Variations on typical modeling parameters were explored in order to gain a better picture of the average grain size and the effectiveness of various measurement techniques. We find that it is possible to use the magnetization measurements to determine the average grain size even if the SEM image analysis allows extracting more information about the grain size distribution. Furthermore a Matlab routine has been developed in order to get automatic analysis of SEM images.