As CMOS technology scales to nanometer dimensions, its performance and behavior become less predictable. Reliability studies for nanocircuits and systems become important when the circuit�s outputs are affected by its sensitive noisy inputs. In conventional circuits, the impact of the inputs on reliability can be observed by the deterministic input patterns. However, in nanoscale circuits, the inputs behave probabilistically. The Bayesian networks technique is used to compute the reliability of a circuit in conjunction with the Monte Carlo simulations approach which is applied to model the probabilistic inputs and ultimately to determine sensitive inputs and worst-case input combinations. © Springer International Publishing Switzerland 2015.