Mechanical imperfections are almost inevitable in large-span spatial domes with hundreds of members and joints. The lack of fit in members is the most common form of imperfections in double-layer domes which influences the behavior of these structures under gravitational or seismic loads. In this study, the probabilistic effect of mechanical imperfections (lack of fit) on the stability behavior of double-layer Scallop domes with sinusoidal arching style under symmetric and asymmetric snow-loading is investigated. On some double-layer scallop domes, nonlinear static and dynamic analyses are carried out using Abaqus/CAE and MATLAB. The behavior of the structures with randomly distributed imperfect members is compared with those of the ideally perfect structures. A Monte Carlo simulation approach is used to statistically assess the effects of uncertainties on the response parameters and the ultimate capacity of the structures. In the first approach, named here as the “rigorous probabilistic approach”, the amount and the distribution of imperfect members are all treated as random variables, and in the second approach, named as “concise approach”, a fixed amount of imperfections are applied to a number of randomly selected critical members. No simplification method is used in the rigorous probabilistic approach to achieve a reliable assessment of the effect of uncertain parameters. Results show that the collapse behavior of double-layer Scallop domes is sensitive to random imperfections and a considerable decrease is seen in their load-bearing capacity.