This study proposes an innovative methodology by combining experimental testing and numerical modelling for self-evaluating the crack development of conductive cementitious materials based on piezoresistivity. Cube specimens enhanced with carbon black nanoparticles were subjected to cyclic and uniaxial compression loads individually for evaluating their piezoresistivity behaviour and analysing the crack sensing mechanism. A discrete element model (DEM) was utilised to extract crack development information which was correlated with resistivity variation in experiments for damage quantification. Acoustic emission was also utilised for the validation of resistivity-based crack monitoring. It was found that electrical resistivity started to rise at a relatively early stage of compression process, followed by an increasing rise before the specimen reached its peak strength and then a rapid surge when the specimen started to lose its load-bearing capacity. The results prove that the proposed approach can monitor crack development of conductive cementitious materials from a micro scale through resistivity variation and more precisely describe the damage accumulation. The study also demonstrates that DEM is capable of capturing failure mechanism of cementitious materials.