This paper proposes a methodology framework for the reliability assessment of smart steel Moment Resisting Frame structures (MRFs) equipped with Nickle Titanium Shape Memory Alloy (NiTi SMA) connections subjected to blast loading. The reliability assessment framework is formulated based on a two-step approach algorithm. In the 1st Step, the Monte Carlo Latin Hypercube Sampling Strategy simulation (MC-LHS) is adapted to generate the uncertain parameters sample points. Considering the numerical simulations, the 2nd Step employs simplified performance functions and the generated random outcomes from the 1st Step. The proposed reliability approach is verified against direct Monte Carlo Simulation (MCS) and First-Order Reliability Method (FORM). The performance functions are columns’ axial force and bending moments, rotation capacity at the connections, and Inter-Story Drift Ratio (ISDR). Throughout the development of the reliability assessment, the probabilistic models are parametrized on geometrical properties, material properties, vertical loads, model errors, and charge weights. The developed reliability framework is applied to a prototype 4-story smart MRF. The structural safety level is obtained in terms of the Reliability Index (β). The results show that the reliability framework provides an accurate and efficient structural collapse prediction of the MRFs equipped with NiTi SMA-based connections. Finally, sensitivity analysis is performed to indicate the sensitivity of building collapse to blast wave characteristics, material strength, vertical gravity loads, and column profile dimensions. The sensitivity analysis results also confirm the efficiency of the proposed reliability framework in observing the highly sensitive parameters, which is explosive charge weight.